Root Cause Analysis Of Failures Research Articles

Application of natural language processing for spill reduction in an exploration and production company

Spill reduction is a key focus for exploration and production (E&P) companies to protect people, the environment, and to maintain oil production. Learning from previous incidents enable companies to identify ways to lower the severity and frequency of spills. Root cause failure analysis (RCFA) is a key element in process safety management, and this is often focused on high severity incidents because of the time required to properly conduct the analysis. Incidents that have lower severity occur more frequently, but typically less time and resources are spent analyzing these events. Learning from lower severity, high frequency events can be useful in improving process safety programs and reduce spill frequency and impacts because the data set is more representative of the big picture. Natural language processing (NLP) is applied to identify causes and contributing factors of spills and leaks associated with E&P activity to reduce the frequency and severity of these events. Reviewing incident reports allow us to identify trends and areas of improvement, and NLP can be used to process the data quickly and improve data quality. Incident reports are typically free text that includes a description and cause of the event. One challenge with free-text descriptions is that further analysis and interpretation is needed in order to effectively analyze, trend, and learn from the available data. In addition, identifying the causes and contributing factors of a spill accurately can be time intensive. Two natural language processing methods, rule-based entity extraction and a machine learning model approach have been utilized to analyze a greater number of incident reports, which enables us to pinpoint causes of leaks, reduce the frequency of spills, and mitigate the impact of these events. This paper presents the challenges associated with learning from incident reports, particularly spills, and offers solutions that natural language processing can provide to improve data quality and contribute to spill reduction efforts in the context of an exploration and production company.

Failure Analysis of Fuel System Main Engine Fishing Vessel (Case Study : KM. Sumber Mutiara)

Failure analysis on the fishing vessel's main engine is one of the works that must be completed because it is crucial for the operation of the ship. Analysis of determining the cause and effect of the failure main engine fuel system is the goal of this study. FMEA (Failure Mode and Effect Analysis) is a method used to determine priority components in failure. The use of Pareto diagrams is an additional method of analysis. RCFA (Root Cause Failure Analysis) is used to analyze the root cause of the failure of the main engine fuel system. The results obtained are the injector component (168) is the component that has the highest RPN value and is the largest contributor to failure according to the Pareto diagram. The root of the failure identified is the lack of checking before operating. The results of the recommendations obtained are the use of corrective maintenance based on the RPN value and the use of a clean filter for the fuel system. The injector is a component that needs to be maintained so as not to experience a fuel system failure on the main engine. The impact of the results of this study is that it becomes a scientific finding using FMEA and RCFA as a method of analyzing failures in the main engine fuel system on fishing vessels

Comparative Study of Interpretable Image Classification Models

Explainable models in machine learning are increas- ingly popular due to the interpretability-favoring architectural features that help human understanding and interpretation of the decisions made by the model. Although using this type of model – similarly to “robustification” – might degrade prediction accuracy, a better understanding of decisions can greatly aid in the root cause analysis of failures of complex models, like deep neural networks. In this work, we experimentally compare three self-explainable image classification models on two datasets – MNIST and BDD100K –, briefly describing their operation and highlighting their characteristics. We evaluate the backbone models to be able to observe the level of deterioration of the prediction accuracy due to the interpretable module introduced, if any. To improve one of the models studied, we propose modifications to the loss function for learning and suggest a framework for automatic assessment of interpretability by examining the linear separability of the prototypes obtained.

Liquefied natural gas plant maintenance: a case study of gas turbine centrifugal compressor thrust bearing failure analysis

Maintenance is essential for preserving gas turbine engineering system functionality during equipment life cycle. Heavy-duty gas turbine (HDGT) engineering systems are often quite complex to maintain. Therefore an effective condition based maintenance (CBM) program is vital to reduce downtime, improve equipment reliability and cost optimization. It is estimated that 50 to 75% of gas turbine maintenance cost is related to the replacement and repair of hot section parts. Furthermore, the maintenance intervals of the engine is governed by the hot section material durability, and the replacement of these components is the largest contributor to determining maintenance related equipment availability. This paper investigate multiple thrust bearing failures on gas turbine driven centrifugal compressor application in a liquefied natural gas (LNG) plant. A root cause failure analysis (RCFA) approach was conducted to determine the failure mode and its consequencies. Process instability is believed to be the cause of several thrust bearing failures. Downtime analysis is presented to estimate the the loss of production (LPO).

Damage Analysis of Propeller Blade P/N R815505 with Root Cause Failure Analysis Method on ATR 72 Aircraft

The propeller blade is a component that is the main source of thrust on the ATR 72 aircraft. Based on data from the Aircraft Flight Maintenance Log (AFML) for the period 2019 – 2021 ATR 72 aircraft, the propeller blade component was damaged causing the need for unscheduled removal which had impact losses to airlines such as delays and aircraft-on-ground (AOG). This study aims to find the root causes of damage to the propeller blade components. Analysis was carried out using the Root-Cause Failure Analysis (RCFA) method using fault tree analysis (FTA) and fishbone diagrams as tools so that factors related to propeller blade damage will be found. The results of analysis using fault tree analysis (FTA) and fishbone found that the root cause of damage to the propeller blade was in the area of ​​the heating element (de-icer) as the top event, while the propeller blade operation, propeller blade maintenance and the ATR 72 aircraft operating environment were the intermediate events

Technology Focus: Artificial Lift (March 2022)

What is the one thing you can do that will provide the greatest benefit to your artificial lift operations? For me, the answer to that question is simple: focus on surveillance. Surveillance is the foundation of any artificial lift optimization program. Effective surveillance practices allow you to keep track of how your system is performing against predefined targets, understand the current state of your artificial lift system, and identify which wells are at risk of failure. It also helps you identify ways to prevent failures through better selection, design, and application of artificial lift technology. At the heart of any effective surveillance program is a set of thoughtfully selected key performance indicators (KPIs). Establishing surveillance KPIs helps you set goals and track your progress against them so that you can zero in on the areas in greatest need of improvement. Typical KPIs include production volumes, operational expenditures, downtime, capital expenditures, failure frequency, and intervention frequency. A variety of tools and techniques are used daily to monitor and control the performance of artificial lift systems. As technology has evolved, so have these surveillance practices. Historically, artificial lift surveillance was performed using manual techniques, such as measuring and evaluating produced volumes, collecting acoustic fluid levels, evaluating two-pen charts and amp charts, collecting and analyzing dynamometer cards, collecting pressure and temperature surveys, performing teardowns and root-cause failure analysis, and using well and network models to understand system performance better. As automation technology developed, these techniques were enhanced through real-time data collection, the use of permanent downhole gauges, and real-time control using wellsite intelligence. With the introduction of automation, exception-based management techniques were introduced to leverage the time and energy of personnel to the greatest benefit. This allowed operators to focus their efforts on dealing with “bad actors” and finding “low-hanging fruit” rather than spending their time looking at every single well each day. Today, artificial lift surveillance practices continue to evolve through data-science and data-analytics techniques. Data analytics allow operators to put an ever-growing volume of data to work. Such techniques, generally classified as descriptive analytics, diagnostic analytics, predictive analytics, and prescriptive analytics, often combine statistical and physics-based principles to find opportunities faster than traditional practices can. Descriptive analytics allow operators to answer questions about what has happened, diagnostic analytics help operators answer questions about why the events happened, predictive analytics help operators answer questions about what will happen, and prescriptive analytics help operators answer questions about what should be done. This month’s JPT features examples of three such systems. Paper SPE 207290 provides an example of how automated well models are being used to improve the performance of progressing cavity pumps in a large Middle Eastern asset. Paper SPE 200825 shows how one operator is using artificial neural networks to predict failures in sucker rod pumping applications. Finally, paper IPTC 20254 shows how genetic algorithms can be used to accelerate the surveillance and optimization of a large population of gas lift wells. As an added bonus, one of our recommended papers for further reading, paper SPE 207550, illustrates how data analytics techniques are being used to predict failures in electric submersible pumping applications. Recommended additional reading at OnePetro: www.onepetro.org. SPE 208136 - Gas-Lift-Valves Replacement in Norway Field Using Riserless Light Well Intervention Vessel by Antonio Fiameni, Eni, et al. SPE 207550 - Real-Time Implementation of ESP Predictive Analytics—Toward Value Realization From Data Science by Antonio Andrade Marin, Petroleum Development Oman, et al. URTEC 208406 - Statistical Process Control for Early Detection of Progressive Cavity Pump Failures in Vertical Unconventional Gas Wells by Suren Indrajith Rathnayake, The University of Queensland, et al.

Visual Exploration of Anomalies in Cyclic Time Series Data with Matrix and Glyph Representations

The digitalization of manufacturing involves machines equipped with sensors that collect, produce, and exchange data machine-to-machine and machine-to-human in real-time. As the data generated within a production process can be massive and overwhelming for human users, support is needed to understand and explore this data, and drive decisions from it. First, the data has to be monitored and recorded using methods that can handle massive datasets. Next, the collected data has to be analyzed (often in real-time) to, e.g., (i) identify undetected process correlations, (ii) forecast the product quality, and (iii) perform root-cause analysis of failures or problems. The analysis becomes even more valuable when the production process is divided into repeating tasks, producing a vast amount of comparable data. For instance, in automotive durability tests, engineers investigate an engine's condition using multiple sensors, recording data from repeating test cycles. Tests can span dozens or hundreds of cycles, and thousands of runtime hours, making it difficult for engineers to collect and monitor each iteration's data to detect interesting data, such as anomalies. We propose an interactive visual analytics approach that displays the iterations of durability tests as a collection of color-encoded cycle glyphs to tackle this issue. With our approach, domain users including test engineers can readily monitor the test, detect potential anomalies, and intuitively analyze, report and document the detected anomalies. This research is conducted in close collaboration with our partner from the automotive sector and shows the effectiveness and efficiency of a prototype with a pair analytics evaluation study. We open up directions for future work, including a visual interactive labeling concept for anomaly classification.

ANALISIS PENGARUH KELURUSAN POROS KOPLING POMPA HIDRAULIK TERHADAP VARIASI TEKANAN PADA PLTA MANINJAU

Vibration due to misalignment in the hydraulic pump is detected as the main problem of damage to the pump. Resolved a problem with the alignment of the pump, shaft alignment as an action in overcoming the problem. Vibration measurements are carried out during alignment and misalignment conditions by using a vibration measuring instrument, namely vibration analysis fluke 810. Voltage, current, electrical power, motor & pump rotation, and other operating parameters are observed. Vibration analysis is done by observing the velocity and frequency of the wave signal for alignment and misalignment conditions. Vibration analysis is performed using a comparison method to the ISO 10816-3 tester standard. Finally obtained an ideal vibration in the operation of hydraulic pumps based on the above standards.The method used to analyze using the tool RCFA (root cause failure analysis) which is an investigation process to determine the main causes of a failure mode. From the analysis, we concluded that the formulation RCFA rise broken of coupling.

Root cause analysis of failure of bolts in the low pressure section of a gas turbine in an oil and gas production plant

This research establishes the cause of a severe failure in bolts in the low pressure (LP) section of a gas turbine in an oil and gas production plant. The failure in bolts was reported together with massive failures of turbine blades in the second stage in the hot section of the turbine. The failure was categorized as an unexpected premature failure, given that it was reported to take place after 20 k of service and that the expected service life for this turbine is beyond 100 k. This research had a two-fold purpose; i) to analyse the cause of failure in both components (failed bolts and blades) and ii) to come up with a scenario for the chain of events. Standard microscopic examination techniques (optical metallography and electron microscopy) were used to correlate microstructure of failed components with possible failure mechanisms. Results showed a clear corrosion-fatigue-induced failure in blades. Hardly is there any evidence of fatigue failure in bolts. Bolts appeared to have experienced an excessive load and have consequently been heavily deformed and broken. The failure scenario was proposed, based on the service history, observations, and microstructural examinations.

Failure investigation of stainless steel anodes used in gold electrowinning

Failure of austenitic stainless steel anodes in a gold electrowinning process (EW) was reported after a short period of service (3–6 weeks). Different analytical and characterization techniques were employed to conduct root cause failure analysis (RCFA) of the failed anodes. Electrochemical behaviour of the stainless steel alloy in the industrially supplied and synthesized electrowinning solutions were evaluated using different electrochemical test methods. The main failure mechanism of these anodes was found to be transpassive dissolution. Local solution acidification in the pores and crevices also contributed to the high corrosion rate of the stainless steel anodes as the transpassive dissolution rate increases when the pH of the electrowinning solution decreases. Gold and silver were found in the anodic layer along with typical stainless steel corrosion products (iron and chromium oxides), contributing to gold loss in the electrowinning process. Materials selection, design change, solution and process modifications are recommended to prevent similar failures or maximize the lifetime of the anodes.

Experimental studies of: Laminate composition, drill bit wear out, and chloride ion concentration as factors affecting CAF formation rate

In industrial, aerospace, medical and other mission critical electronics employing high density interconnection PCBs, high DC voltage gradient and adverse environmental conditions facilitate Conductive Anodic Filament (CAF) formation. To prevent catastrophic failures caused by CAF short circuits, monitoring of manufacturing process and proper laminate selection is essential. The purpose of presented work was to evaluate CAF resistance level of FR4, low CTE, CAF resistant and BT/epoxy-based laminates. Test coupons comprising 90 potential hole-hole failure points with 250 μm spacing were fabricated and subjected to accelerated CAF test for 600 h in 85 °C, 87% RH. Surface insulation resistance measurements were performed to monitor coupons' condition. Time to failure of examined samples, microscopic inspection results and failure root-cause analysis are presented.

Technology Focus: Artificial Lift (July 2018)

Technology Focus One of the more exciting aspects of artificial lift is the constant influx of new technologies and new ideas that the discipline invites. Inventors appear to be regularly developing “the next new thing” in artificial lift. Some ideas are so brilliant and so obvious, you ask yourself, “Why didn’t I think of that?” Other ideas seem so zany and so ill-conceived, you ask yourself, “What were they thinking?” In every case, these new technologies have the same goal: To produce more oil and gas more reliably and more cost-effectively. One of my artificial-lift mentors has such a long history of developing and championing new downhole pumping technologies that he once described his career as the “relentless pursuit of the everlasting pump.” Even in retirement, he continues this lifelong passion. Yet, it occurs to me that the everlasting pump may actually exist today—through the correct application of field-proven technology. Although we seem to have a never-ending supply of new artificial-lift technologies, the technologies that consistently allow us to achieve our technical and business goals are usually the old ones. There is a catch, however: You need to apply those technologies correctly. Applying artificial-lift technology correctly means selecting the right lift method in the first place, selecting the right system components, installing the equipment correctly, performing effective surveillance, and performing root-cause failure analysis so that we can learn from our mistakes and enable continuous improvement. All of this requires a coordinated effort, not just between vendors and operators but also between the various groups within the operating companies themselves. Each of these groups has a stake in the success of the well and should be aligned around a common set of goals to ensure that the artificial-lift system performs reliably and delivers optimal production and does so in a cost-effective manner. All of the papers selected for this section concern the correct application of field-proven artificial-lift technology. Paper SPE 181216 describes the surveillance methods used by one operator to improve the reliability and performance of their beam pumping systems. Paper SPE 186254 describes a method to deploy electrical submersible pumps in gassy horizontal wells to improve gas handling and mitigate slugging. Paper SPE 186110, written by one of our recent Legends of Artificial Lift inductees, provides new insights into the application of injection-pressure-operated gas-lift valves, to allow operators to improve the design and overall reliability of gas-lift installations. To see more great artificial-lift papers like these and to welcome our latest class of Legends, I invite you to join us for the 2018 SPE Artificial Lift Conference and Exhibition—Americas on 28–30 August 2018 in The Woodlands, Texas, USA. Recommended additional reading at OnePetro: www.onepetro.org. OTC 28123 A Step Change in Safety and Quality for ESP Deployment in the North Sea by L. Pastre, Schlumberger, et al. SPE 190090 Detecting Failures and Optimizing Performance in Artificial Lift Using Machine Learning Models by Mike Pennel, OspreyData, et al. SPE 187490 A Novel Approach to Distributed Field-Based Gas-Lift Facilities by Siddharth Mullick, Anadarko Petroleum, et al.

Root cause analysis of failures and quality deviations in manufacturing using machine learning

Today root causes of failures and quality deviations in manufacturing are usually identified using existing on-site expert knowledge about causal relationships between process steps and the nature of failures and deviations. Automatization of identification and back tracking of root causes for said failures and deviations would benefit companies both in that knowledge can be transferred between factories and that knowledge will be preserved for future use. We propose a machine learning framework using Bayesian networks to model the causal relationships between manufacturing stages using expert knowledge, and demonstrate the usefulness of the framework on two simulated manufacturing processes.

On the importance of combining “the new” with “the old” – One important prerequisite for maintenance in Industry 4.0

Development and technological advancement within Industry 4.0 is on top of agendas worldwide. The prediction that cost of technology will decrease while computational power will increase may lead to a maintenance community putting too much trust in technological advancement, leaving basic maintenance concepts and management underdeveloped or even unimplemented in the future. This paper will, through a case study, exemplify the importance of combining basic maintenance concepts and management, e.g., early equipment management, predetermined preventive maintenance, and root-cause failure analysis and elimination, with technological advancement, e.g., predictive maintenance, cyber-physical systems, internet of things, and big data, in order to increase total maintenance effectiveness.

Investigation on the mechanical stresses in a muffler mounting bracket using Root Cause Failure Analysis (RCFA), finite element analysis and experimental validation

Abstract In the present work, experimental investigation on the failure of a muffler mounting bracket attached to commercial vehicle is done. Cracks are identified at the welded location of muffler mount which shows that weld joint has better strength than the muffler/bracket body. To understand the possible root causes of the failure, fishbone diagram was used, which helped in determining the major causes of the failure by a graphical representation. Further, the three parameter Weibull distribution was also developed to determine the Mean Time to Failure (MTTF) life which was found to be 15,172 km. In addition, tensile testing of sheet metal was performed on the sheets which was used in the manufacturing of Muffler. Furthermore, a Thermo-Mechanical coupled analysis was carried out using commercial code, ANSYS 16.0, which adapts Finite Element Analysis (FEA) formulation. The thermal loads on muffler were imported to structural analysis along with a static load of 4 g acceleration were imposed on the muffler body to simulate the effects of high impact loads. The effect of temperatures resulted in high von-misses stresses which were in agreement with practical observations. It is observed that the stresses at bracket region were at the same location both from the experimental and finite element analysis. Further, by considering different mounting design concepts, multiple analyses were carried out to determine the most suitable brackets for the muffler. The present work attempts to diagnose the cause of failure and to make the fail-safe design.

Friction-Load Redistribution for Extended-Reach- and Horizontal-Well Completions

This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper IADC/SPE 178905, “Friction-Load Redistribution for Robust and Cost-Efficient Design of Extended-Reach- and Horizontal-Well Completions,” by A.R. McSpadden, SPE, and R. Trevisan, SPE, Altus Well Experts, prepared for the 2016 IADC/SPE Drilling Conference and Exhibition, Fort Worth, Texas, USA, 1–3 March. In horizontal and extended-reach wells in which long completions are run into highly deviated or lateral zones, large compression loads arise because of running friction. These loads remain locked in the string when the packer or cement sets. Dissipation of friction caused by string vibrations and movements redistributes these friction loads between the wellhead and the packer or the top of the cement. A numerical approach is presented to calculate the redistributed friction load so that an accurate initial tubing load is implemented in the tubular-stress analysis. Introduction The long lateral wells in modern shale developments provide new motivation to better understand and model the effect of friction in casing and tubing design. The multiple-barrier approach and modern design techniques compare favorably with previous generations of well construction. Even so, the root-cause failure analysis of individual subsurface components such as tubing, connections, or packers is difficult. The multizone- fracture-stimulation techniques common to shale developments combine the multiple challenges of long lateral zones with associated friction-induced compression loads and cyclic thermal loading from successive stimulation, flowback, and production. An important motivation for understanding the effect of redistribution of locked-in friction loads is the increasing application of onetrip multistage completions and intelligent-well-system completions in unconventional- reservoir field developments. In extended-reach wells, as the long laterals grow longer and more-expensive equipment is run out farther into the reservoir-contact area, design requirements will become increasingly challenging. It can be argued that these considerations are not as important for cemented completions or where the production casing is the effective operational conduit. However, even for production-casing fracture stimulations or even production, loads below the top of the cement cannot necessarily be disregarded. Also, the planned top-of-cement depth may not be realized. Hence, much more of the casing may be uncemented than in the original plan.

Root-Cause-Failure Analysis as a Tool for Investigating Operational Failures: A Case Study

Summary Production-liner cementation in deepwater wells is critical to contain formation pressures and provide zonal isolation. This is especially challenging in the Gulf of Mexico (GOM) because of depleted zones and small drilling margins. Failure to achieve zonal isolation with the primary-cement job leads to added cost and time for remedial operations. A recent primary-cement-job failure in the GOM prompted the use of an established tool, root-cause-failure analysis (RCFA), in a novel way to identify the causes of the incident and avoid repeating the same mistakes in future wells. A team of operations and engineering personnel was assembled to investigate the cause of the primary-cement-job failure that resulted in nonproductive time and cost of approximately 7 days. RCFAs have typically been performed on equipment failures; however, this RCFA was conducted by the operator on a nonequipment-related failure. The RCFA team initially collected data available from both the planning and operation phases of the job to identify which elements of the cementing process contributed to the cement-job failure. After identifying the root cause, the team communicated their learnings to office and rig-based personnel; developed a new tool for real-time monitoring of primary cementation; and adjusted plans for upcoming primary-cement jobs. The findings from the RCFA were applied to a well on another asset owned by the operator, with which the production liner was successfully rotated throughout cementation for the first time. The RCFA identified that the cement-job design was not a contributing factor; instead, the root causes were operational and started during the circulation of drilling fluid, before the cementing operations. The learnings from the RCFA were transferred to both office and rig-based personnel, influencing engineering and operations plans for subsequent wells. Furthermore, the investigation identified previously unknown issues and showed that the use of the RCFA process was an effective approach that can be standardized and scaled to determine root causes of future operational failures.

Run-Life Improvement by Implementation of Artificial-Lift-Systems Failure Classification

This article, written by Special Publications Editor Adam Wilson, contains highlights of paper SPE 173913, “Run-Life Improvement by Implementation of Artificial-Lift- Systems Failure Classification and Root-Cause Failure Classification,” by Edward Rubiano, José Luis Martin, Jesus Prada, Manuel Monroy, Laura Labrador, Jonathan Celis, and Jahir Gutierrez, Oxy, and Michel Bohorquez, Ecopetrol, prepared for the 2015 SPE Artificial Lift Conference—Latin America and Caribbean, Salvador, Brazil, 27–28 May. The paper has not been peer reviewed. This paper describes a methodology for classification of artificial-lift-system (ALS) failures and addition of a commonly used root-cause failure classification. This methodology was applied to different ALSs such as beam pumps (BPs), progressing cavity pumps (PCPs), electrical submersible pumps (ESPs), and electrical submersible progressing cavity pumps (ESPCPs). The starting point was the definition of the boundaries of each system. Then, each job in the well was defined as (1) a failure, (2) a failure, non-ALS, or (3) no failure, in order to look for ways to improve run life. Introduction La Cira-Infantas oil field is in the Middle Magdalena Valley in Colombia (Fig. 1). The vast majority of the wells are completed with 7-in. production casing and perforated with 5 shots/ft with no sand-control completion in place, and then the ALS is installed. ALS in La Cira-Infantas From the beginning of the activity in La Cira-Infantas, almost 100 years ago, several ALSs have been used to produce fluids from the wells. Since 2005, the field has offered different conditions, depending on the mature state of the waterflooding— high water percentage, high and low flow rates, free gas, and corrosive fluids. Currently, ALSs such as BPs, PCPs, ESPs, and ESPCPs are used for producing the total fluid of the field. Failure Analysis The root-cause failure-analysis (RCFA) process used in La Cira-Infantas has the objective of finding the real cause of a given failure and identifying effective solutions to prevent a recurrence of the failure. The steps for the RCFA are as follows. Initial Diagnosis and Failure Confirmation. When an ALS ceases to lift the expected production, specific procedures are performed to try to restart the system. If production is not re-established, a downhole failure is confirmed. Pulling Equipment From the Well. Mechanical and electrical tests are performed to determine ALS condition. Tubing is tested for leaks. As equipment is pulled, each part of the ALS is inspected carefully and qualitative measurements are taken.

Root-cause failure analysis of photocurrent loss in polythiophene:fullerene-based inverted solar cells.

Metal oxide transport layers have played a crucial role in recent progress in organic photovoltaic (OPV) device stability. Here, we measure the stability of inverted and encapsulated polythiophene:fullerene cells with MoO3/Ag/Al composite anode in operational conditions combining solar radiation and 65 °C. Performance loss of over 50% in the first 100 h of the aging is dominated by a drop in the short-circuit current (Jsc). We reveal a concurrent loss in reflectance from 85% to 50% above 650 nm, which is below the optical gap of the used photoactive materials, hence, excluding any major degradation in the bulk of this layer. Correlating the responses of aged devices to a series of test structures comprised of ITO/ZnO cathode, MoO3/Ag, and MoO3/Ag/Al anodes and their combinations with the active layer allowed us to identify that the presence of Al causes the reduced reflectance in these devices, independent of the presence of the active layer. Systematic single-stress aging on the test structures further indicates that elevated heat is the cause of the reflectance loss. Cross-section transmission electron microscopy coupled with elemental analysis revealed the unsuspected role of Al; notably, it diffuses through the entire 150 nm thick Ag layer and accumulates at the MoO3/Ag interface. Moreover, XRD analysis of the aged MoO3/Ag/Al anode indicates the formation of Ag2Al alloy. Depth profiling with X-ray photoelectron spectroscopy advanced our understanding by confirming the formation of Ag-Al intermetallic alloy and the presence of oxidized Al only at the MoO3/Ag interface suggesting a concomitant reduction of MoO3 to most probably MoO2. This latter compound is less reflective than MoO3, which can explain the reduced reflectance in aged devices as proven by optical simulations. On the basis of these results, we could estimate that 20% of the loss in Jsc is ascribed to reduction of MoO3 triggered by its direct contact with Al.

Applying Root-Cause Failure Analysis Enhances Artificial-Lift Performance

This article, written by Special Publications Editor Adam Wilson, contains highlights of paper IPTC 17230, ’Artificial-Lift-Performance Enhancements by Applying Root-Cause Failure Analysis,’ by S.G. Lapi, SPE, and M.E. Johnson, SPE, ExxonMobil Production Company, and B. Arisman, SPE, GE Oil & Gas, prepared for the 2014 International Petroleum Technology Conference, Doha, Qatar, 20-22 January. The paper has not been peer reviewed. This paper describes the use of a root-cause-failure-analysis (RCFA) process to improve artificial-lift- system performance in a project in Chad. The process was established with the objective of evaluating every failed pump system to determine the reason for failure, identify contributing factors, and monitor trends. The process has helped to reduce the failure frequency by more than 70% on electrical submersible pumps (ESPs) and by more than 50% on progressing-cavity pumps (PCPs), despite the fact that the ESP population has more than tripled and PCP installations have more than doubled. Introduction The Chad Project is in the central African country of Chad, approximately 430 km south of N’Djamena, the capital city. The vast majority of the wells are completed with 9?-in. production casing and perforated with 7-in. conveyed perforating guns at 12 shots/ft. The wells are surged to remove perforating debris and are gravel packed for sand control before the artificial-lift system is installed. Typically, ESPs are run with 4½-in. production tubing and PCP systems are conveyed with 5½-in. tubing (Fig. 1). Formation properties are quite favorable, with permeability ranging from 0.5 to 10 darcies and porosity ranging from 24 to 28%. The hydrocarbon-bearing deposits are found at depths as shallow as 3,000 ft and as deep as 6,000 ft. The crude produced from these reservoirs has a gravity of 17 to 24°API, in addition to being heavily biodegraded and undersaturated, with viscosities ranging between 40 and 800 cp. Production is sweet, with no hydrogen sulfide or carbon dioxide, and reservoir temperature is low at 140°F.