Day 2 Wed, September 04, 2019最新文献

{"title":"From Qualitative to Quantitative – How Visual Data Analytics has Transformed Downhole Video Diagnostic Services","authors":"T. Tymons, Glyn Roberts, Christopher Scott","doi":"10.2118/195797-MS","DOIUrl":"https://doi.org/10.2118/195797-MS","url":null,"abstract":"\u0000 Video images have traditionally provided intuitive visual analysis in a wide range of wellbore diagnostic situations. Step changes in computer vision techniques and image processing have led to the ability to make measurements from images (visual analytics). This paper demonstrates several applications where the application of this new data analytics source, combined with state-of-the-art acquisition technology, have further improved understanding of complex well issues while reducing operational time, risk and cost. Examples include hydraulic fracturing, well integrity, erosion, restrictions and leaks.\u0000 The paper will describe the methods and process of this visual analytics technique through discussion of the three main work flow stages from data acquisition to final analytical product, including the innovative developments in sensor, system and computer vision applications that support each step:Acquisition of full circumferential, depth-synchronized video data of the wellbore. An array of four orthogonally positioned cameras, pointing directly at the pipe wall, concurrently record overlapping images, enabling a continuous full-well video dataset to be obtained.The four depth-matched video streams are synchronized and \"stitched\" together through the application of computer vision algorithms to provide a continuous 360° map of the wellbore with submillimeter pixel density.Calibration and measurement of the acquired images before new and unique diagnostic enhancing data analysis methods are applied.\u0000 The paper will provide real-world examples, presented as case studies, for applications including well integrity evaluation, screen condition assessment, and analysis of perforations. Each case study will demonstrate how visual data analytics used to quantify downhole features, combined with the ability to capture a complete, high definition view of the pipe wall, provides detailed and highly intuitive information that leads to an enhanced understanding of the well and the factors affecting its performance.\u0000 We will demonstrate that the application of this visual analytic method, together with the latest generation imaging system, exceeds the limits of conventional logging technologies for multiple industry challenges, such as:Hydraulic fracturing: confirmation and measurement of uniformity of proppant placement to cluster-level and perforation-levelWell integrity: identify corrosion or erosion events, assess their severity and quantify changes with respect to timeWellbore restrictions: understanding of obstructions and their root cause, time-lapse quantification of their extent and progress made with their removal.\u0000 The paper will demonstrate that data analytics when applied to images from the latest generation of downhole video imaging systems has enabled the development of new diagnostics methods that provide unique insight on high value operational issues. This step change in information empowers decision-making leading to improved economics a","PeriodicalId":113290,"journal":{"name":"Day 2 Wed, September 04, 2019","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129776906","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Advances in Automated Design Approval","authors":"Timothy S. Hare","doi":"10.2118/195785-MS","DOIUrl":"https://doi.org/10.2118/195785-MS","url":null,"abstract":"\u0000 There are significant trends in the reduction of traditional 2D design, this is being replaced by the sole development of 3D models. This paper will detail how to develop algorithms to automate large aspects of a design review. These techniques significantly increase efficiency, ensure constancy and optimise the accuracy of the design, leading to reduced project costs.\u0000 Utilising the 3D models enriched metadata and by developing independent algorithms, it is possible to create a cyberphysical model that enables automation of the design review. For example; using the geometrical data in the 3D model to check a hazard with respect to a detector, confirming that the detector is located close to the hazard. There are multiple checks similar this example, cataloguing and scripting these checks can be managed within PLM software.\u0000 Using algorithmic automation techniques reduces the overall design hours of a project, it checks the consistency of the design. Getting it right first time reduces the number of changes later in the project lifecycle, avoiding expensive rework costs. During the first phase of this initiative, we have found, that automation leads to a reduction of design hours by 10% and increases the accuracy and consistency of the design review.\u0000 This first phase of automation uses the metadata in the 3D model, where the output from the check leads to a comment on the design. To scale the pilotm which will encompass the inclusion of other data sources, will further enrich the cyberphysical model. Ultimately, by creating a decisions database and using Artificial Intelligence we will be able to close the loop, which will lead to a design that is fully evaluated before it leaves the designer. It is also possible to automate in other phases of the project lifecycle, where image recognition will compare the real asset to the model.\u0000 This level of automation is unique, there are other low-level forms of automation, but the advancements of this technology has, to our knowledge, not been attempted in the Oil and Gas sector. The development and scaling of this technology is novel and will have a significant impact on the way future projects are executed.","PeriodicalId":113290,"journal":{"name":"Day 2 Wed, September 04, 2019","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134496504","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Evaluation of Directly Simulated WAG Hysteresis at Pore Scale and its Effect on Injectivity Index","authors":"A. Fager, B. Crouse, Guangyuan Sun, R. Xu, D. Freed","doi":"10.2118/195734-MS","DOIUrl":"https://doi.org/10.2118/195734-MS","url":null,"abstract":"\u0000 Water Alternating Gas (WAG) injection is a widely practiced EOR method for many reservoirs. One drawback of WAG is the decreased injectivity when gas, often CO2, is injected into a previously water-flooded reservoir, and a further decline of injectivity is observed as water and gas injection are alternated. We present a workflow which allows the estimation of injectivity decline using pore scale displacement simulations and reservoir simulations.\u0000 In this approach, we use a multiphase Lattice Boltzmann method to directly simulate the alternating water-gas injection at pore scale resulting in a relative permeability curve for each injection phase. The simulation input accounts for injection rate, fluid properties and spatially varying wettability for each cycle during WAG. The final distribution of fluid phases in pore space of each displacement test is used as the starting point for the next displacement cycle. This enables the simulation of imbibition-drainage cycles. Any hysteresis effects present are typically captured in the resulting relative permeability curves. These are then used in a reservoir model to obtain an injectivity index for each injection phase.\u0000 We observe a strong decline of water relative permeability after the first gas injection cycle in an oil-wet rock. Detailed analysis of the fluid phases, in particular the water phase, shows that water is well connected after the initial water flood before gas injection. As gas is injected large water blobs are partially displaced and their size significantly reduced. For this wettability scenario, water and gas are competing for the large pore system. We find that capturing the hysteresis effect in a WAG requires the direct simulation of the displacement process, in particular known pore scale phenomena such as trapping and retraction.\u0000 The novelty of this approach is to directly capture the hysteresis effect of a WAG workflow in a direct simulation of displacement at pore scale. Emphasis is put on a detailed analysis of the multiphase displacement, including visualizations and an explanation for why the injectivity during WAG is reduced, namely, water and gas are competing for the same pore space. The presented workflow enables an a priori estimate for injectivity losses in a WAG EOR approach.","PeriodicalId":113290,"journal":{"name":"Day 2 Wed, September 04, 2019","volume":"77 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115995310","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Digital Worker: Empowering Offshore Operators with Easily Accessible Data to Improve Efficiency and Safety","authors":"A. Aadland, C. Straumsheim, A. Giset","doi":"10.2118/195749-MS","DOIUrl":"https://doi.org/10.2118/195749-MS","url":null,"abstract":"\u0000 Digitalization is the transformation of business models and activities through the strategic use of digital technologies. Despite technological advancements in machine learning (ML), artificial intelligence (AI), and virtual reality (VR), there remains a low maturity of digitalization across the oil and gas industry, especially in offshore operations. There are many roadblocks on the way to digitalization, from data silos to legacy systems. Operational inefficiency is one of the most painful byproducts of these problems.\u0000 To complete a single maintenance task, for example, on-site workers may need to access several separate systems to get the required data. They rely on printing out the information they need in order to complete the maintenance activities, and after taking notes on pieces of paper, they have to return to their desktop computer to log the performed tasks.\u0000 Not having the data readily accessible contributes to overall inefficiency, and offshore workers often run back and forth while performing maintenance tasks, increasing the hours they spend in challenging conditions.\u0000 This paper will outline an application design philosophy for oil and gas companies that combines academic and practical insights, an emphasis on continually testing products in development, and an overall goal of creating value.\u0000 This paper will describe how a Nordic software company is using the design philosophy to help an oil and gas operator in Northern Europe optimize on-site operations -- including increasing efficiency and safety -- on its offshore installations on the Norwegian Continental Shelf.\u0000 Specifically, the paper will show the software company ingested and contextualized operational data from the operator's assets and made historical data available for field workers via an application for computers and smart devices. This included access to sensor data and historic equipment performance data; all documentation related to maintenance, including procedures, drawings, piping and instrumentation diagrams (P&IDs), and maintenance logs; and interactive 3D models of installations and equipment.\u0000 After only three months, the crew at one of the operator's oil installations saw significant increases in the number of monthly maintenance jobs (up to 10% for certain tasks) and reduction of the time spent on certain routine inspections (in some cases up to 50%).","PeriodicalId":113290,"journal":{"name":"Day 2 Wed, September 04, 2019","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129364170","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Decarbonizing Anthropogenic Activity: The Oil and Gas Industry is a Major Component of the Solution","authors":"A. Hastings, Pete Smith","doi":"10.2118/195716-MS","DOIUrl":"https://doi.org/10.2118/195716-MS","url":null,"abstract":"\u0000 The major challenge facing society in the 21st century is to improve the quality of life for all citizens in an egalitarian way, by providing sufficient food, shelter, energy and other resources for a healthy meaningful life, whilst at the same time decarbonizing anthropogenic activity to provide a safe global climate. This means limiting the temperature rise to below 2°C. Currently, spreading wealth and health across the globe is dependent on growing the GDP of all countries. This is driven by the use of energy, which until recently has mostly derived from fossil fuel, though a number of countries have shown a decoupling of GDP growth and greenhouse gas emissions from the energy sector through rapid increases in low carbon energy generation. Nevertheless, as low carbon energy technologies are implemented over the coming decades, fossil fuels will continue to have a vital role in providing energy to drive the global economy. Considering the current level of energy consumption and projected implementation rates of low carbon energy production, a considerable quantity of fossil fuels will still be used, and to avoid emissions of GHG, carbon capture and storage (CCS) on an industrial scale will be required. In addition, the IPCC estimate that large scale GHG removal from the atmosphere is required using technologies such as Bioenergy CCS to achieve climate safety. In this paper we estimate the amount of carbon dioxide that will have to be captured and stored, the storage volume and infrastructure required if we are to achieve both the energy consumption and GHG emission goals. By reference to the UK we conclude that the oil and gas production industry alone has the geological and engineering expertise and global reach to find the geological storage structures and build the facilities, pipelines and wells required. Here we consider why and how oil and gas companies will need to morph into hydrocarbon production and carbon dioxide storage enterprises, and thus be economically sustainable businesses in the long term, by diversifying in and developing this new industry.","PeriodicalId":113290,"journal":{"name":"Day 2 Wed, September 04, 2019","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124444211","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Digitalization's Role in Shaping the New Energy Landscape","authors":"H. Devold, Tor-Eivind Moen","doi":"10.2118/195764-MS","DOIUrl":"https://doi.org/10.2118/195764-MS","url":null,"abstract":"The realization that fossil fuels are a limited resource, and the growing awareness of the negative impact their emissions have on the planet, has impacted every oil and gas major. The global challenge is expressed in the \"energy trilemma\" of: Enough Energy, Affordable Energy and Sustainable Energy. The industry must adapt, in terms of cost and environmental footprint. In this paper we discuss how digitalization and renewable sources can drive innovation to meet these challenges. We will use current long-range forecasts to understand how the global energy mix is expected to change over time, and illustrate how different scenarios are likely to affect the offshore industry. We also study how digitalization and hybridization with technologies like offshore wind and power-from shore, can reduce costs, energy consumption and emissions. There are many trends accelerating the introduction of new energy sources These include: Global population growth and changing dynamics: \"Millennials\" bring with them their own expectations about technology, the pace of work and accountability. Equally influential, is the challenge to feed and power the 2 billion poorest and the extra 2 billion people expected by 2050.Transportation changes: Road, aviation and shipping account for more than 60 percent of the world's oil consumption and key to limiting the impact on the climate.Energy generation revolution: The grid needs to cope with the increased power demands and to incorporate and expand the contribution of renewablesRise in distributed generation: Hybridization pilot projects to use offshore wind turbines to power e.g. water injection systems. There are a range of technologies described, which will provide the transformational step change to enable companies to transition into the broader energy ecosystem. However, the real game changer lies in integrating these technologies in a way that drives the evolution from connected operations, to collaborative operations and ultimately autonomous operations to achieve maximum value. We will describe how, by properly using digital technologies, the sector can not only reduce capital and operating expenditures by up to 30 percent but also use energy optimization and hybridization with renewable energy sources to reduce emissions and help oil and gas operators do their part in addressing \"The Energy Trilemma\".","PeriodicalId":113290,"journal":{"name":"Day 2 Wed, September 04, 2019","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131925095","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The Clair Ridge Project - A Safety Leadership Journey to Create a Strong Culture of Care","authors":"C. White, Rob Houston, G. Chapman, S. Gourlay, K. Lang","doi":"10.2118/195782-MS","DOIUrl":"https://doi.org/10.2118/195782-MS","url":null,"abstract":"\u0000 The Hook-up and Commissioning program for the BP operated Clair Ridge facility was conducted over a period of three years, starting with the accommodation platform in 2015/16, and then the Production and drilling platform over 2017 and 2018. The total topsides weight is 53,000 tonnes, and the field is located in the harsh waters of the Atlantic West of Shetland. Typically 750 persons were based offshore, but over the life of the program some 7000 individuals worked offshore at some point on the project. Recognizing the safety leadership challenges with such a major hook-up and changing workforce a huge amount of effort went into preparation and working with our contractors to onboard the workforce. Over the first months of the campaign the safety metrics were healthy and there was a good reporting culture, however an increase in incidents was seen, including one late in 2015 where a medical evacuation was required from the platform. The individual made a full recovery and returned to work however it caused the Operator and Contractor project leaders to reflect on their safety leadership and how they were working with and engaging with the workforce. It was a catalyst for change as the team was determined that no other serious incidents would happen during the project delivery.\u0000 In this paper we will share the Clair Ridge safety leadership journey and the steps taken by the operator, with the support and collaboration of the main contractors, to set a new approach to safety through the development of a genuine Culture of Care. This included: Building of trust and credibility between leadership and the workforceLeadership openness and transparency in communicationEmpowering front-line supervision to be safety leaders and giving them the skills and tools to do this well\u0000 As a result of the approach the Clair Ridge team is proud that, in the three years since the incident in 2015, over 9 million offshore workhours have been completed without any other Lost Time Incident, and a safe start-up was achieved with no process safety related incidents. Clair Ridge realised some of the highest participation in safety observations and near miss reporting across the Operator's global projects portfolio, a continual and significant reduction in all injuries and benefited from an excellent reporting culture.\u0000 A Culture of Care has been owned by all, and been recognised and commended by the contractor workforce and visitors to Clair Ridge.","PeriodicalId":113290,"journal":{"name":"Day 2 Wed, September 04, 2019","volume":"70 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115859835","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Realtime Lubricating Oil Analysis to Predict Equipment Failure","authors":"O. Alabi, Robert J. Wilson, Urenna V. Adegbotolu, Surakat Kudehinbu, S. Bowden","doi":"10.2118/195708-MS","DOIUrl":"https://doi.org/10.2118/195708-MS","url":null,"abstract":"\u0000 Oil condition monitoring for rotating and reciprocating equipment has typically been laboratory based. A technician or engineer collects a sample of lubricating oil and sends this to a laboratory for chemical analysis. After the laboratory has performed the analysis the results are sent to the engineer to make decisions on the health and/or condition of the machinery. This process can take up to 6 weeks, and consequently analysis may end up being performed only quarterly with little likelihood of critical failures being pre-empted. The slowness of oil condition monitoring analyses performed in laboratories has led engineers to substitute for real-time monitoring methods such as vibration analysis and thermography. Nevertheless, the chemical composition of the lubricating oil remains the gold standard for the diagnosis of machine health. The automation of methods for analysing the chemical composition of lubricating oil in real-time would provide engineers with data on the immediate condition of a particular piece of machinery, allowing the early diagnosis of incipient faults.\u0000 In this paper, we present a microfluidic technique that can perform real-time continuous monitoring of the chemical composition of lubricating fluid from rotating and reciprocating equipment. Results from this technique both in laboratory and field environments are comparable to conventional laboratory measurements. The microfluidic technique exploits the flow of fluids within micrometre-dimensioned channel, permitting liquid-liquid diffusive separation between otherwise miscible non-aqueous fluids. It can be shown that several fluids e.g. methanol, hexane etc. can selectively extract target components in lubricating oil. Following an extraction, these components can be quantified using a combination of optical techniques, e.g. UV/Vis, Infrared etc. This microfluidic technique has been demonstrated for a range of lubricating oils with several acid, alkaline detergent, asphaltene/insoluble content. This technology can potentially revolutionise the way oil analysis is carried out, automating and making the process rapid and in real-time.","PeriodicalId":113290,"journal":{"name":"Day 2 Wed, September 04, 2019","volume":"49 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122100566","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Fractal Approach to the Modelling and Simulation of Heterogeneous and Anisotropic Reservoirs","authors":"P. Glover, P. Lorinczi, Saud Al-Zainaldin, Hassan Al-Ramadhan, Saddam Sinan, G. Daniel","doi":"10.2118/195778-MS","DOIUrl":"https://doi.org/10.2118/195778-MS","url":null,"abstract":"\u0000 New reservoirs are increasingly more heterogeneous and more anisotropic. Unfortunately, conventional reservoir modelling has a resolution of only about 50 m, which means it cannot be used to model heterogeneous and anisotropic reservoirs effectively when such reservoirs exhibit significant inter-well variability at scales less than 50 m. This paper describes a new fractal approach to the modelling and simulation of heterogeneous and anisotropic reservoirs. This approach includes data at all scales such that it can represent the heterogeneity of the reservoir correctly at each scale.\u0000 Three-dimensional Advanced Fractal Reservoir Models (AFRMs) can be generated easily with the appropriate code. This paper will show: (i) how 3D AFRMs can be generated and normalised to represent key petrophysical parameters, (ii) how these models can be used to calculate permeability, synthetic poro-perm cross-plots, water saturation maps and relative permeability curves, (iii) the effect of altering controlled heterogeneity and anisotropy of generic models on fluid production parameters, and (iv) how AFRMs which have been conditioned to represent real reservoirs provide a much better simulated production parameters than the current best technology.\u0000 Results of generic modelling and simulation with AFRMs show how total hydrocarbon production, hydrocarbon production rate, water cut and the time to water breakthrough all depend strongly both on heterogeneity and anisotropy. The results also show that in heterogeneous reservoirs, the best production data is obtained from placing both injectors and producers in the most permeable areas of the reservoir – a result which is at variance with common practice. Modelling with different degrees and directions of anisotropy shows how critical hydrocarbon production data depends on the direction of the anisotropy, and how that changes over the lifetime of the reservoir.\u0000 We have developed a method of fractal interpolation to condition AFRMs to real reservoirs across a wide scale range. Comparison of the hydrocarbon production characteristics of such an approach to a conventional krigging shows a remarkable improvement in the modelling of hydrocarbon production when AFRMs are used; with AFRMs in moderate and high heterogeneity reservoirs returning values always within 5% of the reference case, while the conventional approach often resulted in systematic underestimations of production rate by over 70%.","PeriodicalId":113290,"journal":{"name":"Day 2 Wed, September 04, 2019","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127524498","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Work Orders - Value from Structureless Text in the Era of Digitisation","authors":"Erik Salo, David McMillan, R. Connor","doi":"10.2118/195788-MS","DOIUrl":"https://doi.org/10.2118/195788-MS","url":null,"abstract":"\u0000 Free text and hand-written reports are losing ground to digitization fast, however many hours of effort are still lost across the industry to the manual creation and analysis of these data types. Work orders in particular contain valuable information from failure rates to asset health, but at the same time present operators with such analytical difficulties and lack of structure that many are missing out on the value completely. This research challenges the current mainstream practice of manual work order analysis by presenting a methodology fit for today’s context of efficiency and digitization.\u0000 A prototype text mining software for work order analysis was developed and tested in a user-oriented approach in cooperation with industrial partners. The final prototype combines classical machine learning methods, such as hierarchical clustering, with the operator’s expert knowledge obtained via an active learning approach. A novel distance metric in this context was adapted from information-theoretical research to improve clustering performance.\u0000 Using the prototype tool in a case study with real work order data, analytical effort for certain datasets was reduced by 90% - from two working weeks to a day. In addition, the active learning framework resulted in an approach that end users described as \"practical\" and \"intuitive\" during testing. An in-depth review was also conducted regarding the uncertainty of the results – a key factor for implementation in a decision-making context.\u0000 The outcomes of this work showcase the potential of machine learning to drive the digitization of not only new installations, but also older assets, where as a result the large amount of unstructured historical data becomes an advantage rather than a hindrance. User testing results encourage a wider uptake of machine learning solutions in the industry, and particularly a shift towards more accessible in-house analytical capabilities.","PeriodicalId":113290,"journal":{"name":"Day 2 Wed, September 04, 2019","volume":"120 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-05-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133271227","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}