8 1 0 2 / 3 e u s s I e Journal Pipeline Technology Journal INTEGRITY MANAGEMENT www.pipeline-journal.net ISSN 2196-4300
Advanced in Sealing. DENSOLEN® Tapes and Systems ■ Reliable corrosion prevention and strong mechanical resistance ■ For extreme temperatures: -50 °C to +100 °C (-58 °F to +212 °F) DENSOMAT® Wrapping Devices ■ For fast and secure application ■ High flexibility for all diameters www.denso.de
Safety, Reliability, Profitability – the main drivers of a good integrity management system Pipelines are the veins of a worldwide energy system, serving industry and clients with all kinds of liquids and hazardous fluids and gases. In the 1920s to 1950s they have been built with a commercial lifetime of 25 to 30 years, but today some of the systems are reaching their 100 years of operation an- niversary. The main bulk of high pressure transmission lines in the oil and gas industry will have their 50 or 70 years anniversary. Aging infrastructure per se is no problem for safe operation as long as best maintenance procedures and methods are applied. Today’s successful operation within the oil and gas industry is based on the triangle “Safety - Reliability - Profitability (Efficiency)”. It is of high importance to properly balance these different and sometimes opposite positions. PIPELINE TECHNOLOGY JOURNAL 3 EDITORIAL Heinz Watzka EITEP Senior Advisor High technological and operational standards guarantee safety for human and environment. Innovative technologies ensure security of supply and grant reliability. Competitive service provides efficient transmission conditions for the client and stands for profitability. Different company management systems support the a. m. triangle. A well-advanced Pipeline Integrity System (PIMS) is a major success factor for integrity within technics, organisation and information within the organisation of the transportation system operator (TSO). The relevant standards and codes in Europe (DIN EN 16348) establish the targets of a PIMS with PDCA (Plan-Do-Check-Act) methodologies, relevant documentation, well-defined organizational structures, safety aspect/targets/programs, communication and the development of clear and smart KPIs. Despite its high population density, Germany reaches the highest safety figures (oil and gas transmission systems) compared to international records and publications. Key for this positive trend is the establish- ment of a well-advanced PIMS with regular third-party checks and experience exchange between the differ- ent operators and their associations (DGMK and DVGW). A further increase of safety and therefore decrease of accidents and incidents may be reached with a strong exchange of experience between operators, service companies and manufacturers and furthermore the regulator as the external and responsible part of the a. m. triangle. Long-term commercial cooperation between operators and manufacturers are further possibilities to overcome lack of knowledge due to demographic challenges. It must be a clear and communicated target for the responsible TSOs to establish international platforms for the proposed experience exchange and integration of relevant authorities and regulators. Based on well-ad- vanced and creative communication, this challenge also may improve the missing acceptance of our society (public perception) in respect of pipeline projects and future energy demands. Some few events already take care of this idea but the international pipeline community still has plenty of room for further improvement. This edition of ptj focuses on new developments in intelligent data management from ILI runs, advanced PIMS methods and new sensors and procedures for improving pipeline integrity. All these developments are part of a company’s PIMS and the triangle mentioned above and will support TSOs in keeping their license to operate. Yours, Heinz Watzka, EITEP Senior Advisor, former Technical Director of Open Grid Europe We are working constantly to uphold the continuous exchange within the international pipeline community. Kind- ly find additional information on our websites or contact us directly via mail: • • • email@example.com www.pipeline-journal.net www.pipeline-conference.com
4 PIPELINE TECHNOLOGY JOURNAL THIS ISSUE’S COMPLETE CONTENT AUGUST 2018 / ISSUE 3 TECHNICAL ARTICLES RESEARCH / DEVELOPMENT / TECHNOLOGY Development of a Novel Subsurface Monitoring and Oil Leak Detection System Dr. Stephen Edmondson / Dr. Kaushik Parmar / Adrian Banica Direct-C Steel Pipeline Failure Probability Evaluation Based on In-line Inspection Results Maciej Witek GAZ-SYSTEM Assessing Repeat ILI Data Using Signal-to-Signal Comparison Techniques Jane Dawson/ Geoffrey Hurd Baker Hughes, a GE Company Condition Assessment for Optimizing Gasunie’s Network Improvement Program M. Hommes / K. van Bloemendaal / R. Coster / M. Gielisse / M. van Agteren / E.E.R. Jager DNV GL / Gasunie Transport Services Data-driven Approaches to Pipeline Cleaning Otto Huisman ROSEN Group Trial of a Process for the Identification of Reduced Depth of Cover on Buried Pipelines Daniel Finley / Simon Daniels / Klaas Kole / Michiel Roeleveld / Paul Ogden ROSEN Group / National Grid INDUSTRY NEWS • Creaform and Olympus Announce Worldwide Distribution Agreement for Pipeline Integrity Assessment Solution • GE plans to split from Baker Hughes • Pipeline Transport Institute completes test runs of pipeline transportation hydrodynamic processes test bench • TransCanada has awarded Spiecapag and Macro Pipelines to build two sections for the Coastal GasLink gas pipeline • Pipeline Transport Institute presents energy efficiency benchmarking results • NDT Global Appoints President • BHGE Breaks Ground on European Customer Solutions Center for Inspection Technologies Business • Russian Edition of Pipeline Technology Journal agreed 10 16 22 28 36 42 6 7 8 9 REPORTS CONFERENCES / SEMINARS / EXHIBITIONS www.linkedin.com/ groups/4740567 ptc 2019 Preview www.twitter.com/pipelinejournal ptj Job & Career Market www.facebook.com/ Pipeline.Technology.Conference Event Calendar www.pipeline-journal.net Company Directory Page 54 48 52 57
6 PIPELINE TECHNOLOGY JOURNAL INDUSTRY NEWS Creaform and Olympus Announce Worldwide Distribution Agreement for Pipeline Integrity Assessment Solution Olympus now distributes Creaform’s Pipecheck Analyze software Creaform, a well known company in the business with portable 3D measurement solutions and engineering services, announced has announced that Olympus® Scientific Solutions Americas, a manufacturer of phased array flaw detectors for corrosion inspection, will now distribute Pipecheck™ Analyze, a ophisticated NDT software for pipeline integrity assessment. Pipecheck™ Analyze software solution supports phased array (PA) and conventional ultrasonic testing (UT) data files for corrosion analysis and allows users to gain important information about the status of their components. “We are very proud to partner with Olympus and to in- tegrate OmniScan® data with our code compliant cor- rosion software analysis. The capability of analyzing internal and external corrosion separately or together pushes pipeline assessment to another level,” says Steeves Roy, NDT Product Manager at Creaform. With Pipecheck, NDT service companies and pipeline engineers can get more reliable and traceable analyses to ensure a safe assessment. Pipecheck to this day is the only trusted solution available that enables the identifica- tion of potential issues on both the inner and outer linings of pipes using both ultrasonic testing and 3D scanning, whether they be corrosion, dents or gouges in the metal. Pipecheck can now process data from 3D scanners as well as data from ultrasonic testing devices, such as the Olym- pus’ OmniScan. Pipecheck provides true wall thickness assessment analysis based on the combination of various integrity assessment calculations. Adding Pipecheck to Olympus’ product offering will enable users to use phased array ultrasonic testing ( PAUT) data, and combine that data with the advanced algorithms and strength calculations offered within Pipecheck, to create a very accurate and realistic damage evaluation of pipeline integrity. GE plans to split from Baker Hughes Despite the fact that Baker Hughes just recently joined the General Electric group, the financially struggling American company publicly announced its intention to separate from Baker Hughes in the next two or three years. The reason for this move lies in the intention of GE to focus on its more profitable branches aviation, power and renewable energy. GE will therefore separate from its branches health care and transportation. The decision came after the company has undergone an internal strategic review. Because Baker Hughes has got a two-year-lockup- agreement with General Electric, the split is not likely to be finalized until 2020. The announcement is only partially a surprise, since GE’s CEO John Flannery said in the past, shortly after the merger, that he will seek ways out of the deal. Flannery was appointed CEO in August 2017, long after the merger was a done deal. Baker Hughes remains optimistic about the split. The company has benefits regarding the access to GE’s technolo- gies and a favorable market position, as a company spokesperson stated.
PIPELINE TECHNOLOGY JOURNAL 7 INDUSTRY NEWS Pipeline Transport Institute completes test runs of pipeline transportation hydrodynamic processes test bench The Pipeline Transport Institute has successfully completed factory testing of a self-developed test bench for studying hydrodynamic processes related to pipeline transportation of oil and petroleum products. This test bench, which has a variable profile, will allow experts to study transient pro- cesses in multiphase hydrocarbon flows, including simulating and studying flow of liquid via a gravity flow pipeline (with the possibility of changing the profile of the pipeline) and modelling batching of various hydrocarbon fluids as well as hydraulic shock, gas removal from the pipeline, accu- mulation of water at low points and water removal at different angles of inclination of pipelines. In addition, it will be possible to simulate oil and petroleum products leakage as well as to test methods of detection thereof. The newly developed test bench will provide the necessary conditions for analysing the effectiveness of technolog- ical solutions before their actual implementation at entities of the Transneft system. Transneft and entities of the Transneft system have obtained patents for the technical solutions underlying the test bench. The test bench will be installed at the premises of the Pipeline Transport Institute’s Research and Development Centre in Ufa. Installation will be complete and the bench will be commissioned in Q4 2018. TransCanada has awarded Spiecapag and Macro Pipelines to build two sections for the Coastal GasLink gas pipeline project in Canada A pipeline welder works on an extension of the NGTL System, in Northern Alberta, Canada (Copyright: TransCanada) Canadian energy infrastructure operator TransCanada Corporation has awarded a contract to a joint venture made up of Spiecapag Canada Corp, a VINCI subsid- iary and operational leader, and Macro Pipelines Inc. to build two sections of gas pipeline in the province of Vancouver, British Columbia. The C$900 million (about €585 million) contract includes the construction of more than 166 kilometres of gas pipeline as part of the 670 km Coastal GasLink Pipeline. The contract amount is split, with 60% going to Spiecapag and 40% to Macro Pipelines Inc. The joint venture will carry out a pre-construction planning phase pending a positive final investment decision by LNG Canada* for a proposed natural gas liquefaction facility in Kitimat, British Columbia. The decision is expected in the fourth quarter of 2018, with construction set to get under way in early 2019. “Alongside our joint venture partner Macro Pipelines Inc., we are proud to be part of the Coastal GasLink Pipeline Project and to furnish our expertise in gas pipeline construction. Our ability to perform works in mountainous environ- ments with steep slopes enabled us to win this large contract. Additionally, the project will provide opportunities to qualified local businesses and suppliers along the pipeline route and employment for roughly 900 people hired direct- ly,” said Bruno Guy de Chamisso, Chief Executive Officer of Spiecapag. In October 2017, Spiecapag and its partner Macro Pipelines Inc. also won the contract in British Columbia to build a 36- inch oil pipeline as part of the Trans Mountain Expansion Project.
8 PIPELINE TECHNOLOGY JOURNAL INDUSTRY NEWS Pipeline Transport Institute presents energy efficiency bench- marking results to International Association of Oil Transporters A meeting of the Permanent Expert Group for Energy Effi- ciency of the International Association of Oil Transporters (IAOT) has been held in Prague (the Czech Republic). The meeting brought together representatives of Transneft, The Pipeline Transport Institute (PTI), MERO ChR (the Czech Republic), Transpetrol (Slovakia), Gomeltransneft Druzh- ba (Belarus), MOL (Hungary) , KazTransOil (Kazakhstan), CPC-R, the China National Petroleum Corporation (CNPC) and Ukrtransnafta, that has recently joined the association. Yakov Fridlyand, Director General and Chairman of the expert group for energy efficiency, and Bronislav Grisha, Head of the Energy Efficient Technologies of Oil and Petroleum Prod- ucts Transportation Laboratory, represented PTI at the event. Results of the pipeline transport energy efficiency benchmarking assessment held by PTI in 2017 at 20 process sections of pipelines belonging to the association’s member states were presented at the meeting. The calculations took various technical parameters and properties of the crude oil transported via all the pipelines covered by the study into account. The study results indicated that accomplishment of measures to enhance energy efficiency of crude oil transportation via pipelines of the IAOT member states enabled a 3.5% drop in average specific energy consumption in 2017 versus 2016. The studies also contained recommendations on how to curtail energy consumption further. PTI offered the participants to share the best practices of energy efficiency benchmarking among companies of the oil and gas sector. NDT Global Appoints PresidentNDT Global, supplier of ultrasonic pipeline inspection robotics and integrity services solutions, today announced the appoint-ment of Mr. Richard Matthews as the President of NDT Global.With more than 30 years of experience in the oil and gas indus-try and most recently held the position of Operations Director for PIMS of London, Mr. Matthews’ appointment supports NDT Global’s product strategy and continued growth in developing service solutions to meet the future needs of the industry.“I am both honored and delighted to be the President of NDT Global. I believe our customer-driven research and development focus, along with a commitment to operational rigor and disci-pline, ensures that we continue to offer the best value pipeline assurance solutions in the industry.” Mr. Matthews commented.Based at NDT Global headquarters, he will be responsible for implementing the organization’s strategy and driving the day-to-day business of the company, including the delivery of high-accuracy pipeline robotic solutions for the inspection of cracks, metal loss and mechanical damage to the oil, gas and petrochemical industries worldwide.
PIPELINE TECHNOLOGY JOURNAL 9 INDUSTRY NEWS BHGE Breaks Ground on European Customer Solutions Center for Inspection Technologies Business Baker Hughes, a GE company, has broken ground on a new European Customer Solutions Center (CSC) for its Inspection Technologies (IT) business, one of the world’s leading providers of non-destructive testing (NDT). The CSC will be housed on IT’s existing Wunstorf, Germany site and will be the flagship CSC for European cus- tomers and partners. BHGE will invest a signifi- cant amount in the millions of dollars in the new 9.250 sqm CSC and plans to add up to 100 jobs to the Wunstorf site as part of the project. The announcement follows the grand opening of IT’s largest CSC globally in Cincinnati, USA, earlier this year. Like Cincinnati, the Wunstorf CSC will also bring the most advanced NDT technologies under one roof, including x-ray, CT, ultrasonic, remote visual inspection and sensor solutions. Given the Wunstorf site’s heritage as BHGE’s radiography centre of excellence, the CSC will have a specific focus on 2D X-ray systems and 3D computed tomography (CT), supplemented by high-tech applications for ultrasonic and electromagnetic inspection. In addition, the facility will also house managed services for parts inspec- tion and allow for personalized setups for training and collaboration. “The manufacturing industry is changing, and Industrial Internet of Things coupled with our innovation in X-ray, CT, and other inspection technologies enables us to set new standards in industrial quality and product reliability assurance,” said Holger Laubenthal, CEO of Inspection Technologies for BHGE. “We will support our customers through this change and that’s where a place like this Customer Solutions Center will be a huge asset. Here in Wunstorf, our experts will work together with our customers to develop technologies to solve individual challenges. We like to see ourselves as problem solvers, and I believe no one else in the industry can offer this level of high-quality service for non-destructive testing. “ Russian Edition of agreed Pipeline Technology Journal The international publishers EITEP and Radiofront have agreed to implement a Russian-speaking edition of the Pipeline Tech- nology Journal (ptj). Both companies are committed to meet the high demand for top-class pipeline technology case-studies, technical articles and current industry news. The new Russian edition (“ptj-Вестник трубопроводных технологий”) will be available in Russia, Belarus, Ukraine and Uzbekistan. Its con- tent will be similar to the original ptj but enriched with addition- al content related to the Eastern European pipeline industry. This cooperation enables international pipeline technology and service providers interested in the Russian market to show their know-how and to advertise specifically in an edition distributed in Russia and its neighboring countries. Vice versa, Russian pipeline companies can provide their content and advertisements for publication within the original ptj. For all readers, this cooperation means a better access to insightful technical articles from oil & gas operators and technology & service providers. “The appearance of the Russian version of the ptj is a logical continuation of the EITEP strategy aimed at the creation of a common platform for intense technology exchange between international pipeline operators, service & technology providers.”, said Dr. Klaus Ritter, President of EITEP, who is also well-known for the annual Pipeline Technology Conference (ptc). “Such an exchange will help the global pipeline industry to minimize incidents and to maximize pipeline safety, longevity and profitability”, he added. His counterpart, Aleksey Turbin, General Manager at Radiofront, stated: “Russian companies are keen to be involved in the impres- sive efforts of EITEP to foster the exchange of state-of-the-art-technologies and best-practices. The Russian edition of ptj ist going to be of great use for achieving this worthwhile goal which is in correspondence with the trend of technology globalization”.
Development of a Novel Subsurface Monitoring and Oil Leak Detection System - SubSense LDS Dr. Stephen Edmondson; Dr. Kaushik Parmar; Adrian Banica > Direct-C Abstract Leaks from oil pipelines, storage tank and other facilities can be disruptive, expensive and can cause significant damage to the environment. The conse- quences of such leaks have been well published in recent years, leading to increased political pressure on the industry to find improved ways of monitoring for leaks. According to data published by the Pipeline and Hazardous Materials Safety Administration (PHM- SA), 45 % of oil transportation pipelines in the United States are over 50 years old. More than 600 leaks are reported every year with an annual clean- up cost to industry of over USD550M. To minimize the damage from any leakage, rapid detection of a failure event is essential. Since pipe- lines are usually located in remote areas and buried underground, accomplishing this is often a challenge. Existing leak detection systems are also typically ca- pable of detecting larger leaks more effectively than smaller ones, needing some complementary solution if proper leak monitoring coverage is to be achieved. This article describes a new technology and meth- od for direct hydrocarbon leak detection in the subsoil using a system called SubSense™ LDS. The system consists of Direct-C’s proprietary polymer nanocomposite based hydrocarbon leak detection sensor and a remote communication system. Polymer nanocomposites provide a unique approach to leak detection as they can detect the presence of the smallest amount of hydrocarbon through a change in the electrical properties of the material. This system is particularly well suited for instru- menting high consequence locations such as urban areas, water crossings, and other environmentally sensitive areas with a fast, deterministic and cost effective liquid hydrocarbons detection solution.
POLYMER NANOCOMPOSITE (PNC) COATINGS FOR HYDROCARBON DETECTION The sensor system is comprised of a sensing element consisting of a silicon-based polymer embedded with conductive nanoparticles. This system was developed at the University of Calgary by Dr. Park and Dr. Parmar. The polymer’s characteristic of swelling in the presence of hydrocarbon molecules is exploited. The polymer also provides the advantage of being hydrophobic and thus unaffected by water and ice. Figure 1: Effect of Hydrocarbon Exposure on Polymer Nanocomposite Coating The silicone-based polymer swells upon absorption of hydrocarbon molecules, causing increases in the dis- tances between nanoparticulates thereby increasing the resistance of the silicone-based nanocomposite polymer coating as shown above. Hydrocarbon Pentane Octane Diesel Crude Oil Motor Oil Instantaneous Slope (degrees) 89.3 88.8 73 9 6 Type of Response High High Medium Low Low Table 1: Change in Resistance of the PNC Coating on Exposure to Liquid Hydrocarbons RESEARCH / DEVELOPMENT / TECHNOLOGY PIPELINE TECHNOLOGY JOURNAL 11 Dr. Stephen Edmondson “24/7 monitoring, zero false positives The polymer nanocomposite coating employed in the SubSense™ unit was formulated as to not be affected by methane thus eliminating any false positives caused by naturally occurring methane. Instead, it was tuned to detect C5 to C24 liquid hydrocarbons. The effect of the exposure of this coating to 5 ml of various liquid hydro- carbons is shown in the table and graph above. Since the detection method is based on the rate of change in resistance of the sensor, the type of hydrocarbon can be determined. This enables the discrimination, in a situation where several different hydrocarbons are being stored or transported close to a given sensor, of which particular hy- drocarbon has leaked. This also eliminates the false positive caused by the “wrong” type of hydrocarbon coming into contact with the sensor, for example a spill of diesel fuel onto a sensor would trigger a different sensor response compared to a leak of crude oil from a pipeline onto the same sensor. ADVANTAGES OF NANOCOMPOSITE BASED DETECTION - ELIMINATION OF FALSE POSITIVES A common problem with leak detection systems that use a secondary measurement such as acoustic or flow to detect leaks and infer the presence of a hydrocarbon, are false positives. These false alarms are generated when the presence of a hydrocarbon is falsely inferred due to interferences by other disturbances. For SubSense™ we have analyzed the potential routes to a false positive de- tection and determined if our detection algorithm would generate a false alarm under those conditions. Cause PNC Sensor System False Posi- tive Possible Oil Present Sensor Power Fails Sensor Circuit Breaks High Temperature High Pressure Water Other Chemicals Resistance increases at 6° in 10 s. to over 100 % Signal drops to 0 Resistance goes to ∞ Change of < 20 % in Resistance Resistance decreases No change Resistance increases if polymer swells Shear Force applied Resistance goes to ∞ - No No No No No Yes No Figure 2: Response of the Polymer Nanocomposite Coating Table 2: Possible Routes to a False Positive
12 PIPELINE TECHNOLOGY JOURNAL RESEARCH / DEVELOPMENT / TECHNOLOGY The algorithm used for detecting hydrocarbons is based on an increase in resistance of the coating with a mini- mum 6 degree initial slope and a change in resistance of over 100%. The only identified route to a false positive is exposure to a chemical which would cause swelling at the same rate and magnitude as that given from an exposure of the sensor to hydrocarbons. There are no environmen- tally available chemicals that are known to cause such a response, therefore environmental exposure will not generate false alarms. All other causes described above will give a very different change in the detected signal, which would be measured by the monitoring system and reported, but would not generate an alarm indicating the presence of hydrocarbon. CONFIGURATION OF SUBSENSE™ UNITS The SubSense™ Leak Detection sensor and commu- nication system is primarily targeted at existing liquid hydrocarbon pipelines and storage facilities. The unit is installed in a hydrovac’ed hole next to the pipeline and located in the expected leak path of the hydrocarbons. It features the proprietary Surface Access Port (SAP) in- stallation to allow the hydrocarbon to readily come into contact with the sensing elements. The key performance criteria that are desired for any leak detection systems are listed below: Performance Criteria Reliability Location Detec- tion Accuracy Sensitivity / Scale of Leak Speed / Re- sponse Time Continuous Mon- itoring Direct Detection Effective in Steady-state & transient condi- tions Optimal / Target SubSense Capability > 2 years between servicing + / - 10 meters High High < 5m3 / hour Very High Within a few minutes Continuous moni- toring 24/7 Direct detection, no False Positives Steady-state & transient 1 minute Yes Yes Yes Table 3: Performance Criteria for Leak Detection Systems The SubSense™ unit features four sensors located in a hollow tube at the base of the unit and a communication package at the top of the tube which contains a modem to send out data as shown here. There are a number of communication op- tions, in this instance a cellular modem was used to send a signal out when a hydrocar- bon was detected. A satellite modem or lo- cal radio system could also be employed. TESTING OF SUB- SENSE™ UNITS The objective of the testing program was to demonstrate the opera- tion of a SubSense LDS sensor in a laboratory environment using a setup representative of field conditions. During this testing program, the prototype sensor was placed in a similar soil sample that it would see in the field. The sensor within the test setup was surrounded by gravel inside a porous PVC pipe, similar to field in- stallation methods. The porous PVC pipe was surrounded by sand contained in a clear acrylic tube or stand- pipe to visually observe the oil contamination level. The contaminants (gasoline, diesel fuel and crude oil) were Communi- cation Unit 4 sensors inside tube Figure 3: SubSense™ Unit Description
RESEARCH / DEVELOPMENT / TECHNOLOGY PIPELINE TECHNOLOGY JOURNAL 13 The algorithm employed to send the text mes- sage was the detection of a high initial gradient when the resistance changed on exposure of the sensor to a hydro- carbon, it was not based on the total change in resistance. TEST RESULTS WITH HYDROCARBONS Prior to Gasoline expo- sure testing, a sensor was held submerged overnight in well water (high mineral content) to simulate a flooded or submerged condition. No change in voltage was observed. On exposure to gaso- Figure 4: Test Set Up at C-Core then introduced through a port in the side of the acrylic standpipe. The contaminants flow through the sand and into the sensor tube. line, all three sensors triggered. The response was very rapid due to the relatively short chain hydrocarbons present in the product. The sensor readings were monitored to determine when they changed indicating the presence of a hydrocarbon. The sensor was exposed to three types of hydrocarbon: 1. Gasoline- 87 octane from local gas station 2. Diesel- from local gas station 3. Crude Oil- 857 density All three sensors triggered a text message alert to a cell phone. The algorithm was set to alert for the smaller chain hydrocarbons, so this response was expected. On exposure to Oil, all three sensors triggered as shown n Figure 5 and 6. The response was less rapid than for the other hydrocarbons, as expected, therefore no text message alarm was sent. The reservoir bucket was filled with the contaminant hy- drocarbon fluid. As the fluid level increased, the sensor was monitored to determine if the voltage changed upon contact with the hydrocarbon. The algorithm was set to alarm for smaller chain hydro- carbons, this shows that SubSense has successfully alarmed for smaller chain length hydrocarbons exposure and has not alarmed due to crude oil exposure. A positive response, or trigger, could be indicated either by the sending of a text message from the unit to the op- erator or by the observation of a rise in voltage from the laptop data acquisition system showing the sensor input rise from approximately 1 volt to 2.9 volts. The alarming algorithm can be tailored to enable the alerting of particular types of hydrocarbons and to eliminate false positives due to other chemicals coming into contact with the sensor. The test was complete when either all sensors triggered along the test strip or when the fluid level had passed all of the sensors.
14 PIPELINE TECHNOLOGY JOURNAL RESEARCH / DEVELOPMENT / TECHNOLOGY CONCLUSIONS • • The design and validation testing of the SubSense™ ground probe, a stand-alone, wireless, liquid hydrocar- bon leak detection unit is complete. The SubSense unit must be installed in the leak path of the liquid since it needs direct contact with a hydrocar- bon to trigger. • Different types of wireless commu- • • • • nication can be used within this unit: cellular, satellite and radio. The polymer nano-composite sensor employed in this unit is stable even when the sensor is fully submerged in water for long time. In independent testing, a large response within one minute for every hydrocarbon tested was ob- served as soon as the liquid contact- ed the sensor. The algorithm employed for detecting and alarming an exposure to hydro- carbons eliminates most routes to a false positive. Text alarms or notifications can be programmed so that the algorithm only alerts for certain types of hydrocarbons allowing for selective detection. Authors Adrian Banica Direct-C President & CEO firstname.lastname@example.org Figure 5: Gasoline exposure Figure 6: Crude Oil exposure Dr. Stephen Edmondson Direct-C Chief Scientist email@example.com Dr. Kaushik Parmar Direct-C VP R&D firstname.lastname@example.org
VERSATILE. Always a leading innovator, we supply customers with cutting-edge diagnostic and system integrity solutions. This, bound with our focus on flexibility, reliability, cost and quality, leads to offerings beyond your expectations. www.rosen-group.com
Steel Pipeline Failure Probability Evaluation Based on In-line Inspection Results Maciej Witek > GAZ-SYSTEM Abstract The main goal of this paper is to estimate onshore buried pipeline failure probability based on Magnetic Flux Leakage (MFL) inspection data. Degradation of an underground steel structures during their service life leads to reduction of the pipe wall thickness. Periodic in-line inspections are performed by grid operators to detect corrosion anomalies and size their depth, length and width. In diagnostics of steel pipelines, it is common practice to track the same flaws in different inspections (i.e. so-called defects matching) based on the lon- gitudinal and circumferential positions of the anomalies reported by applied tools. A code-based engineer- ing approach to estimate the failure pressure was selected as appropriate to be applied directly after in-line inspections, due to the scope of the available data, before any expansive field excavations for direct observa- tions. Det Norske Veritas DNV-RP-F-101 analytical method of burst pressure calculation for a straight pipe was applied. A probabilistic methodology was used to evaluate the severity of part-wall external corrosion defects and their growth over time on gas transmission grid. The Monte Carlo numerical method was selected in this paper for estimation of pipeline failure probability due to the external corrosion with respect to statistical distribution of input parameters. The predicted flaw depth growth was modeled as non-linear with a power law function parameters derived from literature [1,6,7]. The ex- pected defect length growth rates was forecasted as linear with several scenarios. It was assumed that failure probability of an underground pipeline is influenced only by the growth of the existing features, whereas gen- eration of new defects is neglected. The paper illustrates reliability-based maintenance planning, in the case when a number of anomalies and its statistical distributions are known from MFL in-line inspection. Criteria and formulation of a limit state function were presented to determine the burst pressure and corresponding failure probability of a pipeline DN 700, X 52 steel grade with amount of 138 fully matched single part-wall defects. The results of this study shall help maintenance engineers to solve the problems of an effective strategy in reliability-based high pressure gas pipelines management.
18 PIPELINE TECHNOLOGY JOURNAL RESEARCH / DEVELOPMENT / TECHNOLOGY the defects depth power low function parameters were assumed in this paper as follows: d T p d mean 0 0 164 . u T 0 78 . (6) A defects length growth rate in axial direction was mod- eled as linear according to equation (4) with following scenarios of corrosion growth rates presented in Figure 2 as a function of time: Scenario 1 (base) - 1.8 mm/year; 1.0 mm/year; Scenario 2 - Scenario 3 - 0.5 mm/year; no defects growth in axial direction. Scenario 4 - The metal losses depths have the same corrosion growth rates values for all the considered scenarios, as shown in Figure 1. The inspections tools biases and random scattering errors as well as probability of defects detection are neglected in the current study. Figure 1: Defects depth growth rate over time forecasted with a power law function RELIABILITY FUNCTION A formula of a limit state function and analytical meth- odology based on DNV-RP-F-101 [5, 8] criteria is applied to determine the failure pressure of a pipeline with a great number of single metal losses. Similar as in pub- lications e.g. [2-9], a pressure difference formulation of a limit state function and Monte Carlo method were ap- plied for the reliability calculations, due to the corrosion without any pipeline extensive excavations and repairs. , in the case of a pipe affected Limit state function g X by a part-wall metal loss, can be expressed as follows: Figure 2a, 2b: Metal losses length growth rates in the axial direction for various scenarios Failure probability for the corroded pipe as a function of time (T) can be expressed as: P g X T , ª ¬ d 0 º ¼ ³ d g X f x T dx i , i 0 (8) P fDNV T where: – failure pressure of the corroded steel pipe as a T fDNV P function of time, [MPa]. The pipeline failure probability resulting from growing corrosion is determined in the current paper with the use of Monte Carlo (MC) simulation [4-9]. For a specific time period, a numerical simulation is conducted by generat- ing random numbers for variables P fDNV max , with respect to statistical distribution of the input parameters specified in Chapter 3. For each evaluation of the limit <0 is counted. state function (7), the occurrence of g X and OP P fDNV OP max g X where: P fDNV – vector of theoretical failure pressures; (7) The failure probability of the whole section of pipeline Pf pipeline(T) at time step T, with the assumption of inde- pendence of individual failures of pipes connected in a series is calculated as a function of time Pft(T) according to formula (9): max – vector of maximum operating pressure of the OP pipeline to be applied.
RESEARCH / DEVELOPMENT / TECHNOLOGY PIPELINE TECHNOLOGY JOURNAL 19 1 P T ft (9) ity study, the number of trials was set as 106, which is enough to ensure the accuracy of probability of failure estimation [2-4]. Computations in the current paper were carried out with Goldsim software. n 1 t 1 P f pipeline T where: P T ft N f N (10) INPUTS DATA EVALUATION FOR RELIABILITY CALCULATIONS Pft(T) – failure probability of individual defects at time step T, [-]; n – number of corrosion anomalies based on in-line inspection data, [-]; N – total number of simulation cycles/trials, [-]; Nf – number of failure events which means simulation cycles when g X <0, [-]. For each external corrosion feature based on the in-line inspection data, the total number of failure events Nf is determined at time step T, after N samples are generat- ed and failure probability of an individual defect can be obtained using equation (10). The smaller the probability of failure, the larger the sample size is needed in Monte Carlo method to ensure the same calculation accuracy. In this pipeline reliabil- For the inputs parameters specified below, the pipe diam- eter and wall thickness are modeled as random variables based on pipe manufacturer certificates. The coefficient of variation (COV) of the random variable [X] equals the ratio between standard deviation StD[X] of the measured values and its mean value. The random variables listed below arise from the real diagnostics results. A flaws size growth rate equal the mean value obtained from the inspections data divided by whole 25 years of the pipe- line service. A detailed analysis of the diagnostics results can be found in publication . The choice of the Gumbel distribution for operating pressure fluctuations in this paper was based on publications [2-5,7,9]. The maximum operating pressure of studied pipeline is MOP 5.5 MPa and standard deviation computed in  from extreme value distribution parameter is equal to s = 0.3. Statistical No. Parameter Uncertainty Coefficients Distribution type Lognormal Lognormal Normal Normal Gumbel Normal Lognormal Parameters n, k fixed/deterministic Fixed/deterministic distributions of all input param- eters for the analyzed pipeline reliability calcula- tions are reported in Table 1. PIPELINE FAILURE PROBABILITY CALCULATIONS A stochastic chart of the studied pipeline failure pressure over time, due to the growth of defects dimen- sions d(T), L(T) for scenario 2 as an example, is shown in Figure 3. It can be ob- served that the burst pressure changes during 60 years of op- erations starting form the second in-line inspection 1. 2. 3. 4. 7. 8. 9. Steel yield strength (fy) MPa Unit Mean value 370.6 Tensile strength (fu) MPa 554.7 Pipe wall thickness (t) mm 11.0 Pipe diameter (D) mm 711.0 5. Maximum operating MPa 5.5 pressure (MOP) 6. Defect depth (d) mm 2.2 Defect length (L) mm 45.1 StD[fy ]= 12.2 COV[fy]= 3.3 % StD[fu] = 19.4 COV[fu]= 3.5 % StD[t] = 0.5 COV[t] = 4.5 % StD[D] = 20.3 COV[D] = 2.8 % s = 0.3 COV[MOP]= 5.5 % StD[d] = 0.6 COV[d] = 26.6 % StD[L] = 34.6 COV[L] = 76.9 % Defect depth growth rate dp(T) as a power law function acc. to equation (5) with pa- rameters n, k Defect length growth rate as a linear function acc. to equation (4) with parameter (cl) mm/yr - - mm/yr 1.8 Scenario 1 (base) - 1.8 mm/year; Scenario 2 -1.0 mm/year; Scenario 3 -0.5 mm/year; Scenario 4 - no defects growth in the axial direction. Table 1: Statistical distribution of input parameters for reliability evaluation Source: Author’s analysis 
20 PIPELINE TECHNOLOGY JOURNAL RESEARCH / DEVELOPMENT / TECHNOLOGY decreases from 17.5 MPa to 14.5 MPa with a chance of 50%. However, there is also a 1% chance that the failure pressure at the start of pipeline operation period will be in the scope of 11÷15 MPa, and at the end of the consid- ered pipeline life cycle period between 5 and 10 MPa, as it can be shown in Figure 3. Figure 3: Stochastic burst pressure of the pipeline over time due to the growth of features dimensions d(T), L(T) for scenario 2 Source: Author’s calculations A burst pressure probability density function for scenar- io 1 at the end of the considered pipeline life cycle period of 60 service years is shown in Figure 4. Figure 5: Failure pressure probability density function for various defects length grow rates corresponding to Figure 2 Source: Author’s calculations The failure probability over a life cycle of 60 years for the features depth and anomalies length considered in this paper are presented in Figures 6 and 7 as well in a logarithmic scale in Figure 8. The calculated failure prob- abilities over 60 years of pipeline maintenance starting from the second inspection, even for non-reinforced defected pipes, are very low and remain lower than a related code-based target value for a so-called normal safety class set in  as not higher than 10-4 per annum. For a high safety class characterised by frequent and intensive human activity in the pipeline souranding area, the target annual failure probability is set as not exeed- ing 10-5 per annum. For the studied pipeline it means that for scenario 1, after 55th year started from the second diagnostics, the most significant defects need to be re- paired due to crossing the target code based probability of failure . For scenario 2, the target failure probability is reached in the 59th year of pipeline operation, as it can be seen from Figures 7 and 8. Figure 4: Burst pressure probability density function for scenario 1 Source: Author’s calculations For the same corrosion velocity in depth, the smaller defect length grow rates assumed for scenarios 2-4 the higher pipeline burst pressure capacities whose distribu- tions are presented in Figure 5. For the same forecasted corrosion in depth, the overall failure probability for scenario 1 has also the highest value compare to the burst probabilities for lower corrosion growth rates in axial direction. Burst pressure change of the pipeline during the service period depends significantly on a defects length growth rate. Computations of failure pres- sure of the studied pipeline showed that the active pipe wall corrosion defects lie within the acceptable values for the foreseen operating conditions characterised by various parameters surveyed in the current paper. Figure 6: Probability of failure over 60-year pipeline maintenance for the defects depth and their length corresponding to the base scenario Source: Author’s calculations
RESEARCH / DEVELOPMENT / TECHNOLOGY PIPELINE TECHNOLOGY JOURNAL 21 operating conditions characterised by various parame- ters survayed in the current paper. The calculated failure probability over 60 years of pipe- line service starting from the second in-line inspection, even for non-repaired defected pipes, are very low and remain lower than a related code-based target value set for a normal safety class as not higher than 10-4 per annum. In the later maintenance years, e.g. after studied pipeline operation life exceeding 50 years a rate of the failure probability increase is strong, which means the rapid aging process of steel underground structure. The employed method is a technique of reliability control and extension of the remaining service life of the cor- roded pipelines. The applied methodology can be helpful for selection of the optimal inspection intervals for steel pipelines to maintain the failure probability within acceptable values as well as can be also used in defects repairs decisions. References . . . . . . . . . Caleyo, F., Velazquez J.C., Valor A., Hallen J.M., Probability distribution of pitting corrosion depth and rate in underground pipeline: A Monte Carlo study, Corrosion Science, Volume 51 (2009), Pages 1925–1934. Zhang S., Zhou W., System reliability of corroding pipelines considering stochastic pro- cess-based models for defect growth and internal pressure, International Journal of Pressure Vessels and Piping, Volume 111-112 (2013), Pages 120–130. Bazan F. A.V., Beck A.T, Stochastic process corrosion growth models for pipeline reliability, Corrosion Science 74 (2013), Pages 50-58. Shu-Xin Li, Shu-Rong Yu, Hai-Long Zeng, Jian-Hua Li, Rui Liang Predicting corrosion remaining life of underground pipelines with a mechanically-based probabilistic model, Journal of Petro- leum Science and Engineering, Volume 65 (2009), Pages 162–166. Witek M., Gas transmission failure probability estimation and defect repairs activities based on in-line inspection data, Engineering Failure Analysis, Volume 70 (2016), Pages 255-272. Sahraoui Y., Chateauneuf A., The effects of spatial variability of the aggressiveness of soil on system reliability of corroding underground pipelines, International Journal of Pressure Vessels and Piping, Volume 146 (2016), Pages 188-197. Tee K.F., Pesinis K., Reliability prediction for corroding natural gas pipelines, Tunnelling and Underground Space Technology, Volume 65 (2017), Pages 91-105. Det Norske Veritas Recommended practice DNV-RP-F101 Corroded pipelines, October 2010. Sikder Hasan, Faisal Khan, Shawn Kenny Probability assessment of burst limit state due to internal corrosion, International Journal of Pressure Vessels and Piping, Volume 89 (2012), Pages 48–58. Authors Maciej Witek GAZ-SYSTEM Manager email@example.com Figure 7: Probability of failure over 60-year pipeline maintenance for the feature depth and anomaly length corresponding to the data in Figure 5 Source: Author’s calculations Figure 8: Logarithmic chat of failure probability over 60-year pipeline mainte- nance for the defect depth and its length corresponding to the data in Figure 5 Source: Author’s calculations CONCLUSIONS Burst pressure of a steel pipeline was calculated in this paper according to DNV-RP-F101 methodology using the real two repeated diagnostic results, without any field excavations for direct assessment. For an underground gas transmission pipeline DN 700 constructed in the year 1986 from steel grade equivalent to X52, the flaws detected with MFL tools were evaluated by means of statistical methods. A burst pressure change of the pipeline during the ser- vice period depends significantly on a metal loss length growth rate as well as on the predicted defect depth in- crease. Computations of failure pressure of the analyzed pipeline showed that the active corrosion defects lie within the acceptable dimensions for the foreseen
Assessing Repeat ILI Data Using Signal-to-Signal Comparison Techniques Jane Dawson; Geoffrey Hurd > Baker Hughes, a GE Company Abstract For pipelines with successive ILI runs the detected population of corrosion defects can be compared to identify both internal and external corrosion growth. Depending on the number of defects to be compared, the assessment can demand significant effort and expertise to ensure accurate and meaningful correlations between often very large ILI data sets. Specialist ILI comparison software facilitates efficient and accurate signal-to-signal matching and the determination of defect specific growth rates across very high defect pop- ulations. However, since ILI as a measuring technique is subject to inherent uncertainties, the prediction of where corrosion is active and the rate of growth from consecutive ILI runs also has a degree of uncertainty. The level of uncertainty is influenced by several sources of error: The ability to accurately match the metal loss sites between the two ILI data sets • Identification of measurement bias associated with the ILI tools • • Understanding the repeatability errors between the two ILI tools There are various approaches that are used to compensate for these inherent errors. For instance, there are different ILI data matching methods that can be used and depending on the level of precision employed and the input data available these will result in varying levels of accuracy. We state signal-to-signal matching is the most precise and accurate approach that can be used over other methods such as box matching, but is there a common understanding of what “signal-to-signal matching” means, what information is required to perform it, what are the ways it can be done and the relative merits? This paper focuses on these questions in relation to comparing magnetic ILI tool data and looks at the challenges for signal matching across magnetic ILI tools with differing resolutions and even from different vendors. In addition, we discuss the importance of understanding tool bias and repeatability and minimizing the impact of these errors.
RESEARCH / DEVELOPMENT / TECHNOLOGY PIPELINE TECHNOLOGY JOURNAL 23 INTRODUCTION Corrosion is considered a major threat to the integ- rity of many onshore and offshore, gas and liquid pipelines. In the presence of water (from either the product or the external environment) unprotected carbon steel will corrode. Corrosion can affect the load carrying capability of a pipeline and, if it contin- ues to grow, it will result in either a leak or rupture release when it reaches critical dimensions for the pipeline. The first line of defence against corrosion damage is by the primary corrosion control systems, e.g., the pipe coating, cathodic protection and by chemical treatments and/or water removal for internal corro- sion. However, with time these primary control sys- tems often deteriorate or fail and the pipeline opera- tor must be able to identify the location and severity of corrosion activity to determine how quickly the integrity of the pipeline is deteriorating. The accurate estimation of the rate of corrosion growth in a pipeline is a key consideration in the development of effective Integrity Management Programs. The determination of the need for, as well as the location and timing of mitigative or preven- tive measures such as CP upgrades, coating repairs, pipe repairs and chemical treatment programs for pipelines carrying corrosive products all depend on assumptions about the rate of corrosion growth. Also, decisions on the re-inspection interval for the pipeline need to consider the remaining life of the un-investigated corrosion defects. Many pipelines have now been inspected using intel- ligent in-line inspection (ILI) tools several times. Us- ing these repeat ILI data sets to determine corrosion growth rates is now an established and recognized best practice with pipeline operators. Depending on the number of defects to be compared, the assess- ment can demand significant effort and expertise to ensure accurate and meaningful correlations between often very large ILI data sets. There are different ILI data matching methods that can be used and depending on the level of precision employed and the input data available these will result in varying levels of accuracy. In the following sections, this paper discusses the challenges associated with the different ILI data matching and comparison meth- ods and the inherent uncertainties in the resulting corrosion growth rates obtained. The paper focusses on the comparison of magnetic flux leakage ILI data. “This Paper considers the challenges for signal matching across magnetic ILI tools with differing resolutions and/or from different vendors. Jane Dawson ILI BASED CORROSION GROWTH RATES Since the general introduction of ILI techniques in the 1980’s and the broad adoption by most operators by the 1990’s/2000’s (for transmission pipelines at least) ILI has become the commonly used method for determining where on a pipeline corrosion is occurring and the dimensions of the corrosion. The advance of technology in this field has resulted in the availability of many types of ILI technology to cater for the large range of pipeline sizes, product types, internal restric- tions, the different forms of pipeline defects that can occur and the ever-present drive to categorize defect types and predict dimensions more accurately. When there is more than one ILI run for estimating the corro- sion growth rates it is now commonplace to compare the two ILI defect populations to estimate the rate of corrosion growth based on defect-to-defect matching. The significant advantage over other methods is that ILI can provide size and growth rate information on the overall detectable defect population giving visibility of what is happening along the entire pipeline. For pipelines with successive ILI runs the detected population of corrosion defects can be compared to identify both internal and external corrosion growth. Depending on the number of defects to be compared, the assessment can demand significant effort and expertise to ensure accurate and meaningful correla- tions between often very large ILI data sets. Special- ist ILI comparison software facilitates efficient and accurate defect-to-defect matching and the depth comparison to determine the defect specific growth between the two runs across the large ILI defect pop- ulations. However, since ILI as a measuring technique is subject to inherent uncertainties, the prediction of corrosion rates from consecutive ILI runs also has a degree of uncertainty. When comparing two sets of ILI data there are two main sources of error [1,2]. Firstly, error introduced due to inaccurate matching of corrosion sites and secondly inaccuracies associated with the growth measurement. The growth measurement error con- sists of two parts; a bias (a systemic difference in
24 PIPELINE TECHNOLOGY JOURNAL RESEARCH / DEVELOPMENT / TECHNOLOGY the prediction of defect depth and is not associated with growth) and a scatter (represented by the tool repeat- ability error). The tool repeatability error can be obtained from repeated measurements of the same set of defects under the same conditions. The effect of bias and the repeatability error can be minimized by using the same ILI tool technology and vendor for both runs (following calibration of the ILI signal data for different levels of magnetization any bias present will be repeatable in both runs and essentially cancels out). There are several matching techniques that can be used; cluster matching, box matching or signal matching . All three methods involve defect-to-defect matching with varying levels of accuracy as discussed below: 2.1 Cluster MatChing and Box MatChing A comparison of the reported clusters in the pipeline listing from each ILI run can be made by aligning the girth weld numbers, relative distances and orientation of the clusters. However, as illustrated in Figure 1, new clusters reported in the second run can make accurate matching of clusters difficult. If the correct clusters are not matched then this leads to errors in the calculation of corrosion growth rates. Also, there are other errors inherent in cluster matching. The reported ILI cluster will be represented by a maximum depth and total axial length even though it could be comprised of multiple corrosion pits of varying dimensions. By comparing the overall cluster dimensions rather than the individual pit dimensions the resulting corrosion growth rate may un- der-estimate the actual growth (see Figure 2). In addition, different ILI vendors report the cluster position differ- ently (some report distance to start of cluster, others to mid-point or max depth position) adding additional com- plexities to align clusters from 2 x ILI’s. Cluster matching is the least favoured method for determining corrosion growth rates from repeat ILI data due to the increased likelihood of data matching errors and lack of precision. The principle of box matching is demonstrated in Figure 2. If the boxes are available from the ILI vendor, then these Figure 1: Example of Cluster Matching Figure 2: Example of Box Matching can be aligned and matched between the two inspections. Box matching removes one of the main errors associat- ed with cluster matching, which is the assumption that the deepest individual corrosion defect is at the same location in both inspection runs. However, in the example in Figure 2, the deepest defect in the old survey is at a dif- ferent location from the deepest defect in the new survey. Therefore, when conducting the calculation of corrosion rates based on cluster matching, the corrosion rate would be under estimated. This figure also illustrates how the boxing and clustering may change between inspections due to sites of corrosion being detected differently or new corrosion occurring between runs. Although box matching allows some of the data match- ing errors associated with cluster matching to be reduced, it can still be difficult to ensure that accurate matches are made between boxes, especially where areas of complex corrosion exist. The box matching approach tends to be used mainly where different ILI vendor data is being compared. In this scenario, the growth error can still be significant as both bias and tool measurement errors are contributing to the overall growth error. 2.2 signal MatChing The Signal matching approach is illustrated in Figure 3. This is proven to be the most accurate method of comparing repeat ILI data sets. Clearly, the more detail that is available the more accurate the matching of the ILI data and hence it is obvious that signal matching will provide the best matching result. Hence, we state signal-to-signal matching is the most precise and accurate approach that can be used over other the methods, i.e., cluster matching and box match- ing, but is there a common understanding of what “signal-to-signal matching” means, what information is required to perform it, what are the ways it can be done and the relative merits?
RESEARCH / DEVELOPMENT / TECHNOLOGY PIPELINE TECHNOLOGY JOURNAL 25 ii. Differences in the ILI tool magnetic field strength (high vs low field) will lead to different accuracies/ detection for lower level defects. iii. Different ILI vendors use different signal conven- tions for representing the ILI signal data, e.g., metal loss/metal gain may be represented as a negative signal/positive signal direction. iv. Different signal modelling approaches e.g., symmetri- cal vs asymmetrical specifications for sizing accuracy. v. Time based vs distance based sampling will lead to differences in the magnetic signal. vi. Thick wall speed effects are highly dependent on length of magnet return path which may be com- pletely different between tools. i.e., difficulties in calibrating the signal data. vii. Repeatability errors are difficult to quantify (as these are usually determined by statistically analysing re- peated pull-through test data) which is not available for vendor 1 vs vendor 2 data. Hence, when comparing ILI data from different vendors it is more difficult to reduce the growth errors to the same extent. Signal matching can still be performed to minimise the data matching errors either manually on selected locations or along the full pipeline using software capable of such precise matching from the dif- ferent data sources. But conducting the detailed signal comparison to minimise the growth error is difficult to perform meaningfully on different ILI vendor data for the reasons discussed above. So, there are clearly two “signal matching” approaches that tend to be used; these differ based on whether the two ILI data sets belong to the same vendor (and same technology) or are different vendor data. These ap- proaches are described in the following table: Different types of Signal Matching Available for the Same and Different ILI Vendor Data Comparisons ILI Vendor X (Previous run) ILI Vendor Y (Previous run) ILI Vendor X (Current run) Signal matching for data alignment. Signal scaling/calibration for depth comparison and growth determination. Repeatability error calcu- lation for growth certainty. Signal matching for alignment and defect matching. Comparison of reported box depths for growth determi- nation. Table 1: Different Types of Signal Matching Available Depending on ILI Data Sources Assuming that similar magnetic flux leakage technologies are being compared. Figure 3: Example of Signal Matching As the name suggests, signal matching involves match- ing the defects via a direct alignment and comparison of the ILI signal data. Hence, the signal data from both ILI runs is required to facilitate the matching process. The signal data can be accessed via the vendor’s proprietary ILI viewing software, however, if different vendor signal data is being compared this can be problematic as different signal conventions are often used by different vendors for representing the ILI data, e.g., metal loss/ metal gain may be represented as a negative signal/ positive signal direction and there is not a common format used. Hence, signal matching should only be attempted by a person trained in ILI signal analysis and is best left to the ILI vendor. Furthermore, as discussed earlier, inaccurate, matching is not the only source of error in the growth assessment process and the inaccuracies associated with the growth estimation itself introduces other errors. The growth error is made up of two parts; a bias associated with the esti- mation of the defect depth and a scatter representing the tool repeatability error. The effect of bias and the repeat- ability error can be minimized by using the same ILI tool technology and vendor for both runs. The direct compari- son of the ILI signals and the use of signal scaling or cali- bration techniques minimise the errors (bias is eliminated and repeatability error is much smaller than the measure- ment error associated with the individual two tool runs). This approach should be referred to as “Signal matching WITH signal calibration” as it has the added step of the detailed signal calibration and comparison to minimise the growth error. This second step is very difficult to perform meaningfully on different ILI vendor data for the following reasons: i. Different ILI vendor tools will have different ILI tool resolution e.g., high accuracy/detection vs low accu- racy/detection for pinholes. These differences can result in false growth calculations.
26 PIPELINE TECHNOLOGY JOURNAL RESEARCH / DEVELOPMENT / TECHNOLOGY When a corrosion growth assessment is being conduct- ed it is important to decide firstly what level of ILI com- parison is possible with the two ILI data sets, what level of accuracy is required. runs (i.e. the error in X2 – X1). It is a result of the non-re- peatable (or independent) portion of the measurement error for the individual runs. We’ve introduced the concept of the growth error in a ILI data comparison, this is described in more detail in the following section. If σ1 and σ2 represent the standard deviations of the non-repeatable portion of the measurement error for X1 and X2, it can be shown that the standard deviation, σ, of the growth error is given by: GROWTH ERROR V 2 V V 2 1 2 (ii) growth rate CalCulation The basic equation used to calculate corrosion growth rate, R, is: R X 2 X 1 / t (i) where X1 and X2 are the corrosion depths at the time of the first and second ILI inspections, and t is the time between inspections. It is noted that this equation cal- culates the average growth rate over the time interval between ILI runs. It does not capture growth rate varia- tions within that time interval, which could result from changes in the conditions that drive corrosion. Each of X1 and X2 is characterized on the basis of a sin- gle measurement, xm1 and xm2.Because of measurement error, the actual values of X1 and X2 given the measure- ment are treated as uncertain (or random) variables. Since R is calculated from X1 and X2, it also is a random variable. Note that an upper case symbol (e.g. X or R) is used to represent a random variable that can assume a range of values, whereas a lower case symbol (e.g. x or r) is used to represent a specific value assumed by the random variable. This is standard probability notation. 3.2 effeCt of growth MeasureMent unCertainty As discussed earlier in this paper, the growth measure- ment error (or repeatability error) has two components: a bias that changes from defect to defect (due to dif- ferences in conditions between defects) and a scatter representing random variations in measurements made under the same conditions. The scatter is represented by the measurement standard deviation that would be obtained from repeated measurements of the same defect under the same conditions. The bias for a given defect equals the difference between the mean of these measurements and the actual size of the defect. The bias is typically uncertain for different defects and is therefore added to the total uncertainty associated with measurement error. The magnitude of both σ1 and σ2 depends on how much of the measurement bias for a given defect is repeatable between the two runs. This in turn depends on a number of factors including ILI tool differences, analysis tech- niques used and also on the types (morphology) of the defects present. Clearly the data repeatability will be low- er and the bias higher when comparing different ILI ven- dor data whereas when comparing the same vendor ILI data, the repeatability will be high and the bias minimal. There are two scenarios to consider when determining the growth error σ : 1. Only scatter contributes to growth error (σ1 and σ2 represent scatter only). This is applicable if the total bias is identical for the two runs, which is rep- resentative if the same vendor data is used in both ILI runs and the analysis method used can identify and eliminating bias. 2. Total measurement error contributes to growth error (σ1 and σ2 represent the total measurement error). This is applicable if the bias is completely in- dependent for the two runs, which is representative if two different tools are used, and conservative if there is partial correlation between the bias values for the two runs. It is also possible to estimate σ directly as the standard deviation of data representing the difference between pairs of measurements of the same defect under condi- tions that are representative of two consecutive ILI runs, e.g., in pull-through testing. In this case, repeatable bias is eliminated by subtracting the two measurements, and σ is estimated directly. It is also possible to estimate σ directly from a direct comparison of the “static” defects present in the two tool runs and calculation of the stan- dard deviation of the differences between the pairs of defect measurements. The “static” defects are defects that would not change between ILI runs (e.g. mill faults and internal metal loss defects in a pipeline transporting a non-corrosive product such as dry natural gas). The term “growth error” is used to represent the uncer- tainty regarding the total growth between the two ILI For example, based on a standard MFL tool resolution where the 80% certainty depth sizing tolerance is ±10%wt the growth error (also at a certainty level of 80%) is:
RESEARCH / DEVELOPMENT / TECHNOLOGY PIPELINE TECHNOLOGY JOURNAL 27 1. ±4.6%wt* when comparing ILI signal data of the same ILI technology and vendor (this value is vali- dated using pull-through testing data), and 2. ±11%wt* when comparing independent ILI signal data from different ILI vendors (equation (ii) is used to calculate this value). * Note that these values are calculated using two-sided normal probability distributions. The growth errors will increase at higher levels of certainty e.g., for 90% and 95% levels of confidence in the growth error and will be lower for higher certainty levels of depth sizing tolerance (when comparing different vendor data). identifying the likelihood of aCtive Corrosion repeatability). The process was developed by incorpo- rating a machine learning approach to a large sample set of both unchanged and grown metal loss anomalies to design a test that can be applied, with a specified certainty, to each identified matched metal loss. This provides an operator with confidence that a growth anomaly being excavated is truly growing and can also be used to reduce unnecessary excavations. ACKNOWLEDGMENTS This paper was first presented at the Pipeline Pigging and Integrity Management conference held in Houston in February 2018 , and is presented here with permis- sion of the conference organisers. The authors would also like to thank Baker Hughes, a GE company for per- mission to publish this paper. When the ILI signal data is comparable (i.e., usually when the same ILI vendor/technology has been used in both runs) a visual comparison of the signal data will provide a first qualitative identification of a change to the metal loss between the ILI’s and/or the occurrence of new sites of metal loss. References . . . . Nessim, M., Dawson, J., Mora R., Hassanien S., “Obtaining Corrosion Growth Rates from Repeat In-Line Inspection Runs and Dealing with the Measurement Uncertainties”. 2008 International Pipeline Conference, Paper No. 64378. Dawson, J., Kariyawasam, S., “Understanding and Accounting for Pipeline Corrosion Growth Rates”. Paper 22 at the 17th PRCI-JTM 11-15th May 2009 Milan. Dawson, S.J., Stanley, L., Kariyawasam, S. and Race, J.M. 2006. Predicting Corrosion Rates for Oil and Gas Pipelines. 2006 International Pipeline Conference, Paper No. 10261. Dawson, J., Hurd, G., ‘Assessing Repeat ILI Data Using Signal-to-Signal Comparison Techniques’. 2018 PPIM Conference The growth error (i.e., repeatability) can be used to define the level of statistical certainty that an observed growth is “real” and is not due to minor differences in detection/ measurement between the two ILI runs. A threshold level can be set using a one-sided normal probability distri- bution and for a selected statistical certainty level (e.g., 80%, 90%, 95%) above which the observed change is deemed to be associated with “active” corrosion. Where a change is below the threshold it does not mean that this is not necessarily active but that we have less cer- tainty that the change is real growth. It is highlighted that the growth error (repeatability) will be different for different ILI tools. It varies based on tool size, tool sensor resolution, type, wall thickness, tool speed and also defect type can have an effect. When comparing different vendor ILI signal data, the repeatability threshold is a much higher value (less desirable), it is calculated statistically using the sizing tolerances of each tool as discussed above (i.e., for different vendor data). In addition to using depth repeatability, a “signal test” has been developed by BHGE that provides addition- al information whether each individual corrosion is deemed to be active (growing) or not. This test com- pares the MFL signal characteristics, from the two inspections completed using defined types of tools, at each matched anomaly to provide a classification of growing or not, independent of the depth (or depth . Authors Jane Dawson Baker Hughes, a GE Company Principal Consultant, Integrity Engineering firstname.lastname@example.org Authors Geoffrey Hurd Baker Hughes, a GE Company Data Analysis Product Support Manager email@example.com
Condition Assessment for Optimizing Gasunie’s Network Improvement Program (GNIP) Martin Hommes; Karen van Bloemendaal; Roelof Coster; Maurice Gielisse > DNV GL Martin van Agteren; E.E.R. Jager > Gasunie Transport Services Abstract The 40 bar regional gas transportation network of Gasunie Transport Services (GTS) consists of, in addition to pipelines, valve stations, pressure regulating and metering stations and gas receiving stations. The majority of these stations have been built in the period 1960-1980. This raises questions on the remaining technical life- time of these stations and adequate measures to comply with safety and transport standards in the future. Gasunie has developed the Gasunie Network Improvement Program (GNIP) in which replacement of these assets is carried out, prioritised on their expected condition. Gasunie use the Deming circle in order to identify lessons learnt from executing GNIP and verification thereof in the GNIP Verification Project (GVP). In the GVP, life-time critical parts of the replaced stations are inspected, in situ as well as in laboratories, in order to assess their actual condition. Lessons learnt and results from the GVP have led to adjustments in the program in terms of scope and pace. DNV GL has supported Gasunie with developing the GVP, has analysed all GVP results and recommended adjustments both for GNIP and GVP. This paper gives first, as an introduction, a general overview of GNIP and GVP. Secondly, the results of the GVP will be presented with a focus on the integrity of valve stations and more specifically the design wall thickness of and the depth of corrosion defects found on D&S piping. Thirdly, the actions Gasunie has taken based on the GVP outcomes will be discussed.
RESEARCH / DEVELOPMENT / TECHNOLOGY PIPELINE TECHNOLOGY JOURNAL 29 INTRODUCTION The Dutch natural gas industry was founded in the sixties of the previous century, upon the discovery of the large Groningen reserve. Since then, Gasunie’s high and medium pressure transmission networks have been extended and adapted to continue meeting changing (market driven) requirements. Therefore, significant parts of the networks are now already approximately 50 years old. Gasunie performed BowTie risk assessments to identify risks related to natural gas transmission pipe- lines. In these risk analyses, integrity, safety and other risks were identified forageing assets that may affect safe and reliable natural gas transmission. GNIP can therefore be seen as a large scale, coherent bundling of preventative maintenance. The set-up and timing of the program are based on a risk-based prior- itization of assets, determining the order in which they are replaced. This way, planning and execution can be continuously monitored and, where possible or required, adapted by for example an increase or decrease of re- placement rate or a change of prioritising order. To this end, the program includes a GNIP Verification Project (GVP), in which the integrity status of removed assets is assessed, thus closing a Plan-Do-Check-Act or Deming circle, as shown in Figure 1. The findings from these BowTie risk assessments result- ed in a decision to initiate the Gasunie Network Improve- ment Program (GNIP) for the regional 40 bar network. In this programme, three types of stations are subject to profound maintenance in the coming 15 to 20 years. A complete renovation is the most significant measure, which in general is the case for valve stations. The three types of stations are: Below ground valve stations; • • Metering and pressure regulating stations; • Gas delivery stations1. The verification project is executed in yearly batches. During the replacement, the assets at the locations are inspected by specialised companies in different stag- es, see section approach for more details. In the final stage of the GVP, all results are collected, analysed and combined with data from the asset database, in order to assess the (integrity) status of the replaced objects. The resulting findings and recommendations are fed back into the risk assessments, into the planning and set-up of GNIP, and back into the set-up and execution of the verification project itself. Figure 1: Plan-Do Check-Act-Circle formed by GNIP and the GVP
30 PIPELINE TECHNOLOGY JOURNAL RESEARCH / DEVELOPMENT / TECHNOLOGY Figure 2: Example of an below ground valve of a valve location in Gasunie’s medium pressure network The condition of the D&S piping is relevant for safety, but after decades of service buried in the ground, the condition of this piping is unknown. Corrosion of D&S piping is the most relevant threat. It may eventually lead to pinholes, causing natural gas to leak and accumulate inside the valve pit, worst case resulting in an explo- sion. In addition to that, possible corrosion may de- crease the strength of a D&S pipe, increasing the risk of breaking it when the valve is operated by field technicians. Both the design or original wall thickness (strength) of a D&S pipe and the occurrence and depth of corrosion, play a role in the safety risk. Firstly, the design wall thick- ness is unknown, nor can it be verified from original This article focuses on the replacement of one of the three types of stations: the valve stations. In the medi- um pressure network, valve stations are used for either sectioning or connecting pipeline routes or for close-in of for example gas delivery stations. These valve sta- tions are typically built below ground. Each valve station consists of several main and bypass valves of differ- ent sizes, makes and models. Each valve is placed in a protective tube, capped with an in-street or in-field cover, providing access for operation and maintenance. Fig- ure 2 shows an example of a below ground valve. Valve functionality, i.e. to prevent flow of natural gas from one end to the other, is also investigated in the GVP, but not discussed in this paper. SCOPE The original BowTie risk assessments revealed that in general, the design of the valve bodies is sufficiently robust and therefore no major integrity problems are foreseen in the bodies for another couple of decades. The most integrity-sensitive items of the valves are the drain and sealant (D&S) piping that are located upon or in- side the valves. There are several types of D&S piping that are used to drain any liquids from the valve, inject either internally or externally a high viscous sealant or grease to the valve seats or inject a lubricant to moving parts. drawings or manufacturer’s specifications. Generally, the wall thickness of the D&S piping is relatively small and it is known that a share of the older valves may have D&S piping with a design wall thickness that is lower than the current minimum required design wall thickness. The year 1990 is important, as at that time, Gasunie’s specifications regarding this aspect were updated. Sec- ondly, the occurrence and amount of corrosion on aged D&S piping is unknown, and these cannot be verified easily in the field. The two main preventive barriers for D&S piping to pre- vent corrosion are the application of protective coat- ing(s) and maintaining cathodic protection (CP). Protec- tive coating may be in perfect condition, but may also have been initially applied insufficiently, may have been damaged at some point in time, may have been needing repair after valve maintenance activities and/or may have deteriorated over the years. It is unknown what the present-day condition of the D&S piping coating is. Fur- thermore, for most valves, it is difficult to collect detailed information available about the exact functioning history of the cathodic protection system over the years. In the GVP project, 146 valves stations have been analysed, comprising of 923 valves and 919 D&S pipes thereof.
RESEARCH / DEVELOPMENT / TECHNOLOGY PIPELINE TECHNOLOGY JOURNAL 31 APPROACH The GVP has four stages, which generate data for ana- lysing the condition of valve stations in general and D&S piping in particular: • welding, clamps, connections and joining to non-car- bon steel materials, mainly stainless steel; General comments for any additional relevant obser- vations as identified by the inspector. • • In-situ visual inspection while the valve station is still operational and the replacement project has not started; In-situ visual inspection when the valves have been excavated and therefore all components can be visu- ally inspected, but have not been removed; • Visual inspection at the company that cleans the • D&S piping and removes the coating; In-shop investigation of the removed and cleaned components at the inspection company, with accu- rate measurements of the wall thickness and the depth of corrosion defects. The results for each valve station include the following: • Valve station, valve number, type of D&S piping, • inspection date and inspector: these are all used for traceability of the results; Condition of each D&S pipe, including coating type and condition, design wall thickness, presence of any corrosion defects, if any, with related depth and location either aboveground or below ground. De aboveground or below ground location is relevant in order to determine if corrosion defects are protected by CP or not. During the inspection, special focus is given to the condition of the potential presence of RESULTS Figure 3 provides examples of the empirical statistical distributions or histograms of the design wall thickness of externally mounted sealant pipes, including the minimum required design wall thickness. A distinction is made be- tween piping originating from valves installed prior to and after 1990. These graphs show that most of the sealant pipes have a design wall thickness exceeding the cur- rent minimum required design wall thickness of 2.5 mm. However, there are a total of 27 observations lower than the minimum required design wall thickness, almost all of them originating from valves that were installed prior to 1990. Based on this observation, valve stations installed before 1991 should be prioritised for replacement. Similar statistical distributions are available for the other types of D&S piping: drain pipes and internal sealant pipes. Figure 4 shows the empirical statistical distribution of the corrosion rate of all external D&S pipes, i.e. drain and external sealant pipes combined. The corrosion rate for each D&S pipe is determined by assuming that corrosion started in the year of installation and that the corrosion rate was constant for the whole life-time. Figure 3: Design wall thickness of sealant piping, installed prior (left) to and after 1990 (right)
32 PIPELINE TECHNOLOGY JOURNAL RESEARCH / DEVELOPMENT / TECHNOLOGY stations can be adjusted based on the reduction on the probability of failure of D&S piping due to the replace- ment of valve stations. The moment the corrosion depth is equal to the wall thickness is defined as ‘pipe failure’ or leakage. This model is based on the probability of failure of D&S piping as a function of age and design wall thickness. Under the fundamental assumption that the corrosion rate and the design wall thickness of the piping are independent of each other, an age-dependent probability of failure of D&S piping can be calculated from the empirical statistical distributions for these parameters. Consecutively, a probability distribution of the time to failure of any pipe can be calculated through a Monte-Carlo procedure. A Monte-Carlo procedure re- peatedly calculates the resulting time to failure for each D&S pipe with a design wall thickness and a corrosion rate that both are randomly drawn from the empirical statistical distributions. Figure 5 shows the resulting probability of failure curves for the sealant piping as function of the year or installa- tion with 10.000 samples2. Figure 4: Probability distribution of corrosion rates A similar statistical distribution is available for inter- nal sealant pipes that are not exposed to the external environment as are external D&S piping and do not have protective coatings nor are protected by CP. A valve station failure is defined as the first failure of one of its D&S pipes. The expected number of station failures over time across the entire Gasunie network can be determined based on the following data from the Gasunie asset register: RISK MODEL A risk model is used to determine the probability of failure of a D&S pipe, i.e. the risk of incidents with loss of containment. The annual replacement rate of valve • • • The number, make, model and year of installation of valves per valve station; The numbers and types of D&S pipes per valve; The probability of failure for each type of D&S pipe. Figure 5: Probability of failure of sealant piping as function of age
RESEARCH / DEVELOPMENT / TECHNOLOGY PIPELINE TECHNOLOGY JOURNAL 33 Figure 6 shows the expected number of corrosion fail- ures at valve stations as function of time assuming the current, constant replacement rate or base-case scenario. Also shown is the curve for the scenario if GNIP were not executed (‘no replacements’). From this figure, it follows that without GNIP, the expected number of station fail- ures in the network is continuously increasing during the simulated period of fifty years. This result underpins Gasunie’s decision to execute the GNIP program for valve stations. At the current replace- ment rate, the expected number of station failures in the network is continuously decreasing, reaching zero the moment all stations are replaced. Similar curves of the number of expected failures over time can be calculated assuming a different number of valve stations replaced each year. Figure 7 shows the expected number of failures in the entire Gasunie network as function of time assuming several other replacement rates relative to the base-case scenario, namely a replacement rate of 33%, 50%, 67%, 80%, 133% and 167% of the base-case rate. These curves show that the replacement rate can be low- ered to a certain extent. When the replacement rate is reduced too much (down to 33% and 50%), the expected number of failures will increase and exceed the current, acceptable level in the first years. A replacement rate of about 67% of the current rate is found to be acceptable while still maintaining the cur- rent number of failures per year. Other simulations were executed, investigating the effect of for example changing the prioritisation order of stations to be replaced. These simulations show that prioritising on age (older stations first) and configuration (certain types of valves first) results in minimizing the number of failures per year thereby minimizing the risk. These findings were taken into account in a re-assess- ment of the required replacement rate within GNIP. As a result, Gasunie decided to lower the rate with 33%. With this decision, significant CAPEX investments are post- poned, while at the same time the replacement process becomes better manageable. Figure 6: Expected number of corrosion failures in valve stations as function of time for the base-case scenario The probability distribution of the time to failure of the valve station follows from the probability distribution of the time to failure of each of its D&S pipes. The expect- ed number of failures across the entire Gasunie network in a given year is the sum of the probability of failure of each individual valve station in that year. Calculating this expected number of failures for all future years re- sults in a curve for the future evolution of the expected number of failures. Figure 7: Expected number of corrosion failures in valve stations as function of time for several scenarios
34 PIPELINE TECHNOLOGY JOURNAL RESEARCH / DEVELOPMENT / TECHNOLOGY CONCLUSIONS RECOMMENDATIONS Based on the work performed, the following conclusions can be drawn: • Results from the GVP confirm the necessity of re- placing valve stations in order to reduce the risk of incidents with loss of containment; Information generated in the GVP is used as input for decisions to adjust GNIP. In particular, for valve stations the replacement rate could be reduced with 33%, while still maintaining an acceptable risk; Information generated in the GVP is used to identify the valve stations that have the highest probability of failure based either on design wall thickness and/ or age. By prioritizing valve stations for replacement based on year of installation and age, the risk reduc- tion can be maximized. • • Several detailed recommendations were identified, most importantly relating to adjustments of the GNIP pro- gram to maximize the risk reduction at the lowest costs. Furthermore, possible improvements of the management and execution of the GVP project, tests and analyses performed were identified in order to improve the quality of data and information generated during the GVP. Footers 1. Recently GTS has decided to stop complete renovation of gas delivery sta- tions and to focus on maintenance. 2. As figure 5 shows, the cumulative probability reaches 0.6 instead of 1. This is caused by the fact that a significant amount of sealant piping hardly cor- rodes, according to the corrosion rate probability distribution (see figure 4). The cumulative probability of failure will therefore not reach 1. Martin Hommes DNV GL Authors Maurice Gielisse DNV GL Principal Consultant Asset Head of Section Asset Integrity Integrity Management firstname.lastname@example.org email@example.com Karen van Bloemendaal DNV GL Principal Consultant Asset Integrity Management karen.vanbloemendaal@ dnvgl.com Roelof Coster DNV GL Technical consultant firstname.lastname@example.org Martin van Agteren Gasunie Transport Services Principle Advisor Asset Management email@example.com E.E.R. Jager Gasunie Transport Services firstname.lastname@example.org
PIPELINE TECHNOLOGY CONFERENCE SEMINARS EUROPE’S LEADING PIPELINE CONFERENCE & EXHIBITION 14TH PIPELINE TECHNOLOGY CONFERENCE 19-21 MARCH 2019, BERLIN, GERMANY 18 MARCH 2019 SEMINAR TOPICS Life-Cycle Extension Leak Detection Inline Inspection Inline Inspection Geohazards Geohazards Corrosion Protection Life-Cycle Extension Leak Detection Offshore Pipelines & Risers Corrosion an event Euro Institute for Information Euro Institute for Information and Technology Transfer BOOK NOW! +49 511 90992-22 Email: email@example.com More Information www.pipeline-conference.com
Data-driven Approaches to Pipeline Cleaning Otto Huisman > ROSEN Group Abstract Data-driven approaches are gaining momentum in the pipeline industry. Proactive pipeline maintenance re- quires the collection and management of data from cleaning programs for future use. This paper illustrates an approach which allows pipeline operators the opportunity to build up a database of information on their assets from standard cleaning runs. Intelligent Gauge Plates and Pipeline Data Loggers (PDLs) can also be integrated in the tool’s setup for more comprehensive analysis. A wide range of analytics can be brought to bear upon these databases. The resulting knowledge of the pipeline conditions offers a greater degree of confidence that a line is ready for further in-line inspection, ultimately increasing first-run success rates while reducing risk.
THE NEED FOR PIPELINE CLEANING The efficient operation of a pipeline is dependent upon maintenance of the internal diameter to ensure optimal flow of the medium. There are a range of significant processes at work inside a pipeline working to decrease flow efficiency. Primarily, ongoing accumulation of de- posits which can either cause damage through abrasion or encourage corrosion as a result of the deposits. Com- promised pipeline surfaces prohibit corrosion inhibitors from being applied consistently. Product contamination can result, and system contamination can complicate the preparatory work necessary to ensure high quality data from an inline inspection (ILI). The absence of a cleaning regime can dramatically affect the efficiency, safety, and reliability of the entire network. Foreign matter and buildup can damage the integrity of a pipeline, encourage the formation of corrosion and pipe thinning, and will almost certainly reduce throughput. As can be seen in Figure 1 below, even smooth deposits can result in a loss of throughput, anywhere between 10- 35% in the case of uneven deposits. Effective cleaning programs are about optimization of the maintenance budget to reduce inefficiencies, max- imizing pipeline uptime and product throughput, and extending the lifespan of the asset. A wide range of cleaning tool technologies exist, includ- ing unidirectional and bi-directional tools ranging from light to heavy duty, and equipped with brushes, sealing and scraper discs and magnets to suit. It is even possible to include speed control options in the newest generation of tools (see Figure 2). Figure 1: Loss of throughput in the case of pipeline deposits RESEARCH / DEVELOPMENT / TECHNOLOGY PIPELINE TECHNOLOGY JOURNAL 37 Figure 2: Speed control to optimize cleaning tool operation THE EMERGENCE OF DATA-DRIVEN APPROACHES Recent years have seen the emergence of data-driven approaches in a wide range of industries. The pipeline industry is no exception. This phenomenon can be attributed to the increasing popularity of approaches such as Risk-Based Inspection and Risk-Based Maintenance Management Frameworks. Data-driven approaches are methods originally devel- oped in the computational sciences in which decisions made are based on the collection and analysis of data rather than pre-conceived ideas or existing knowledge about what is happening in the system. All too often, no data is captured on cleaning run con- ditions with regards to type, volume, or nature of the debris removed during the process. This means that operators may be missing tangible information regard- ing pipeline conditions that could provide guidance on whether an in-line inspection can be conducted smooth- ly, or if the cleaning program is effective. This may result in uncertainties and increased risks for the efficient transportation of products and operational cleaning or inspection tool runs. “Effective cleaning programs are about optimization of the maintenance budget to re- duce inefficiencies, maximizing pipeline uptime and product throughput, and extending the lifespan of the asset. Otto Huisman
38 PIPELINE TECHNOLOGY JOURNAL RESEARCH / DEVELOPMENT / TECHNOLOGY “The systematic collection, organization and management of cleaning data will facili- tate a wide range of analytical approaches. Otto Huisman PiPeline data ColleCtion The concept behind data-driven cleaning is quite sim- ple: collect as much data as possible during launch, cleaning run and receipt of the cleaning pig, and use this information to accurately determine the internal pipeline condition, as well as possible improvements to the tool configuration, tool speed inside the line, inspec- tion interval, etcetera. ROSEN’s new Cleaning Analytic Service was developed specifically to address these issues and provides pipe- line operators with the opportunity to build up a data- base of information on their assets from standard clean- ing runs. The service captures information from multiple sources at the beginning and end of each cleaning run. Figure 3: Ruggedized field tablet with custom form for data collection Implementing such approaches successfully requires the collection of significant amounts of data to drive a robust set of analytics, which in turn provide inputs into planning and maintenance processes. In the pipeline arena, significant complexity is introduced due to the need to combine various design and operational vari- ables, including pipeline diameter, pressure, etcetera. Figure 4: KPI Dashboard presenting summary data and statistics
The Service consists of three main components: 1. Smart Monitoring. Data such as trap conditions, received tool conditions, debris type and volume, and photographic evidence is captured and up- loaded in-field using ruggedized (and if necessary, ATEX certified) hardware such as tablets (Figure 3). Operational cleaning data can also be collected by incorporating intelligent units such as ROSEN Pipeline Data Loggers (PDL) and Intelligent Gauge Plates into standard cleaning tools. Pipeline Data Loggers collect and store operational data during a pipeline inspection. They provide operators with detailed time dependent data such as temperature of the medium, pressure conditions in the pipeline, including differential pressure and acceleration, even indication of bends including bend angle. The latest PDLs can be attached to any standard clean- ing tool and can record for more than 30 days and up to 500 km (310 miles) of inspection. Intelligent gauge plates are a newer technology, developed to assess the internal geometry of the pipeline, able to detect internal deformations which may restrict flow or prevent the passage of an ILI tool. 2. Data Management refers to the transfer of the cap- tured data to a secure ROSEN cloud. Often, Wi-Fi or cellular data connections are available directly in the field. If these are not available, upload can take place as soon as a data connection can be established. From the in-field device, data captured in the form is transferred to a hosted database for analysis and monitoring of the cleaning program. A web dashboard is a key part of the service. This provides a constantly updated view of cleaning operations data (Figure 4) which has been up- loaded. KPIs can be configured in the dashboard to summarize specific data which is collected in the field, providing potentially valuable insights to inform decision making, and if necessary, trigger more detailed investigations into problematic pipe- line locations. Once the data is uploaded It is also possible to draw upon the wealth of other internal databases to better inform the analyses and poten- tial recommendations using a range of data mining and machine learning approaches. 3. Assessment includes the analysis of existing data and detailed reporting. The collective assessment of Operational parameters and monitoring of tool behavior during the run can be utilized to detect RESEARCH / DEVELOPMENT / TECHNOLOGY PIPELINE TECHNOLOGY JOURNAL 39 “The resulting knowledge of the pipeline conditions offers a greater degree of confi- dence that a line is ready for further in-line inspection, ultimately increasing first-run suc- cess rates while reducing risk Otto Huisman and locate restrictions or deposits in the line and provide information on cleaning progress and effectiveness, while verifying operational pipeline conditions. When applied to consecutive runs, such an approach enables the systematic build-up of knowledge about a pipeline’s development over time. A wide range of analytics can be brought to bear upon these databases, including trending of tool disc wear over time, analyses of differential pressure patterns, and ultimately, determination of the optimal way forward with regard to cleaning tool configuration, cleaning interval, and optionally, flow-assurance modelling. It is important to note here that expert knowledge still remains a critical component in the equation, specifically in the area of interpretation of results. Inspection Solutions for Non-Piggable Pipelines World Wide Self propelled BiDi Tethered Inspection Tool Technology is a cost efficient approach. www.ktn.no Norway • Germany • France • Spain • Scotland O f fi c e L o c a t i o n s : KTN NORWAY Postbox 109 Ytre Laksevåg 5848 Bergen NORWAY
40 PIPELINE TECHNOLOGY JOURNAL RESEARCH / DEVELOPMENT / TECHNOLOGY Figure 5: Cleaning tool PDL data analysis showing significant pipeline deformation led to a stuck tool. Buildup of pressure eventually dislodged the tool enabling run completion OUTCOMES The systematic collection, organization and manage- ment of cleaning data will facilitate a wide range of ana- lytical approaches. Photographic evidence, while difficult to employ within an automatic processing chain, can be extremely useful as reference material. The condition of the cups and the disks on the tool and the amount and type of debris that is received can tell much about a pipeline’s current operational status. By analyzing the data from these units, it is possible to verify general pipeline conditions. Specifically, detec- tion and location of restrictions and deposits in the line can be detected by monitoring tool behavior and differential pressure through various data integration and analytical steps (Figure 5). Outcomes of analyses might range from recommen- dations for improving cleaning tool configuration to the identification and assessment of specific problem areas inside the pipeline from detailed PDL analysis. Proactive flow-assurance modelling could be employed for specific cases. Data-driven approaches to pipeline cleaning are a quan- titative approach to pipeline maintenance. The imple- mentation of such approaches requires collection and management of data from cleaning programs for future use. We are currently witnessing a significant move towards automated evaluation of pipeline related data. Analysis and expert interpretation of this data will ulti- mately benefit any additional process by offering more information from the beginning, and potentially decreas- ing the workload of in-line inspections. In a system that can operate in near real-time, status alerts can be provided to give operators critical feedback such as when there is a tool in the line, when a cleaning run was successfully completed, or when a specific prob- lem has occurred. The on-device forms can be config- ured to collect the required information for such notifica- tions. The resulting knowledge of the pipeline conditions offers a greater degree of confidence that a line is ready for further in-line inspection, ultimately increasing first- run success rates while reducing risk. Author Otto Huisman ROSEN Group Product Manager / Sales Manager firstname.lastname@example.org
In the next Edition of ptj: Offshore Technologies The next issue of Pipeline Technology Journal (ptj) will address challenging pipelines. This is a great opportunity for skilled authors to submit insightful papers and to contribute to the global pipeline industry’s constant professional exchange. 14TH PIPELINE TECHNOLOGY CONFERENCE 19 - 21 MARCH 2019, ESTREL CONVENTION CENTER, BERLIN, GERMANY www.pipeline-conference.com Platinum Sponsor Golden Sponsors Silver Sponsors
Trial of a Process for the Identification of Reduced Depth of Cover on Buried Pipelines Daniel Finley; Simon Daniels; Klaas Kole; Michiel Roeleveld > ROSEN Group Paul Ogden > National Grid Abstract Third-party interference is widely documented as being a major cause of damage to buried pipelines. In addition to routine surveillance, maintaining a minimum depth of cover is recognized as a key means of mitigation against third-party interference. We know that the depth of cover over pipelines can change with time. Current techniques available for measuring depth of cover on buried pipes require significant effort to produce a high-resolution sur- vey for an entire pipeline. A UK Innovation project completed for National Grid Gas Transmission has successfully demonstrated a meth- odology to identify reduced depth of cover over an entire pipeline. This methodology combines ground elevation data with high-resolution inertial measurement unit (IMU) data collected during inline inspection to calculate the pipeline depth of cover. GPS and pipe depth measurements have been used to verify the accuracy of this method. Using the pipe center- line derived from the IMU data, and ground elevation data collected using light detection and ranging (LiDAR) techniques, depth of cover has been calculated to an accuracy of ±0.149 m root mean square error. This paper describes the key project steps associated with planning, data collection, data processing, and the vali- dation of results to demonstrate that pipeline depth of cover over an entire pipeline can be accurately determined.
INTRODUCTION Maintaining a minimum depth of cover is recognized as a key means of mitigation against third-party interfer- ence. The United Kingdom Onshore Pipeline Operators’ Association (UKOPA) good practice guide for managing pipelines with reduced depth of cover  states that the best way of determining pipeline depth of cover is to take measurements as part of an over-line survey. The guide recommends measurements should be taken at 50 m intervals but this should be modified depending on topography of the land and any known local issues such as ground erosion. ROSEN Group (ROSEN) and National Grid Gas Trans- mission (NGGT) have trialed a new methodology to identify reduced depth of cover over an entire pipeline. Knowledge of the locations of reduced depth of cov- er can help NGGT reduce the likelihood of third party interference events occurring. METHODOLOGY ground elevation data RESEARCH / DEVELOPMENT / TECHNOLOGY PIPELINE TECHNOLOGY JOURNAL 43 inspection tool on those pipelines which can be moni- tored using such devices. The inspection device typically includes systems to detect corrosion and geometric anomalies such as dents. In addition, inspection devices often include an inertial measurement unit (IMU), these units contain gyroscopes and accelerometers and are used to calculate position of the inspection device. The IMU data can be linked to known reference locations along a pipeline route to provide an accurate pipe cen- terline as a series of X, Y, and Z coordinates. dePth of Cover estiMation The methodology trialed to estimate depth of cover combines ground elevation data with an accurate pipe centerline derived from internal inspection. DEPTH OF COVER REQUIREMENTS In the United Kingdom (UK) guidance on minimum depth of cover for onshore high pressure pipelines is provided in IGEM/TD/1  and PD 8010 . NGGT op- erate their high pressure pipelines in accordance with IGEM/TD/1 Edition 5. Table 1 provides a summary of the minimum depth of cover requirements of IGEM/TD/1 (all editions), PD 8010 and other relevant international standards. IGEM/TD/1 Edition 5 PD 8010-1:2015 ASME B31.8  AS 2885.1 1.1 1.1 1.2 1.2 1.4 0.61 (Class 1) 0.76 (Class 2) 0.76 (Class 3 & 4) 0.91 0.9 1.2 1.2 1.2 1.4 - 1.8 0.91 0.5 0.75 0.9 - 1.2 - 0.9 (W) 0.6 (T1, T2) 0.45 (R1, R2) Location Spec. IGE/TD/1 (Ed. 2, 3 & 4) 1.1 All (m) Rural (m) IGE/TD/1 Edition 1 0.91 (3 ft) Ground elevation data can be collected using several methods. Accurate data for small areas can be collect- ed using differential global positioning system (DGPS) survey equipment. To capture ground elevation data on a larger scale, a LiDAR sensor can be attached to aircraft. LiDAR is a remote sensing method which uses laser light to measure distance to a target and is commonly used to map terrain and sur- face objects. The advantage of this method is that a large amount of highly accurate data can be collected allow- ing large areas to be surveyed efficiently. Watercourses, canals, rivers (m) Suburban (m) Rocky Ground (m) Railways (m) Roads (m) Table 1: Standards Requirements for Minimum Depth of Cover internal insPeCtion Standards for operating high pressure pipelines require that the condition of a pipeline is established periodi- cally. The condition is established by the use of internal Key: R1 – Rural R2 – Rural Residential T1 – Residential T2 – High Density W - Submerged Class 1 – Rural Class 2 – Rural residential Classes 3 & 4 - High density
44 PIPELINE TECHNOLOGY JOURNAL RESEARCH / DEVELOPMENT / TECHNOLOGY DEPTH OF COVER ASSESSMENT The methodology was trialed on a 36” diameter, 45 km pipeline in the UK. PiPe Centerline Following completion of the internal inspection the IMU data was processed to produce an accurate pipe center- line. The output from the processing is a spreadsheet containing a series of X, Y and Z coordinates, Figure 1. Points can subsequently be imported into a geographic information system (GIS) and used to create a pipe cen- terline polyline feature. ground elevation data There were two sources of ground elevation data used within this trial, the Environment Agency (EA) LiDAR and Ordnance Survey Terrain 5 data. The EA  offer LiDAR data with a spatial resolution of between 25 cm and 2 m. It is currently stated by the EA that accurate elevation data is available for over 75% of England. The absolute height error is quoted to be less than ±15 cm. This is the root mean square (RMS) error. The Ordnance Survey (OS) Terrain 5 data  has a quot- ed height error of ±1.5 m. This is the RMS error for urban and major communication routes. For rural and mountain and moorland areas the error is higher at ±2.5 m. The spatial resolution for all OS Terrain 5 data is 5 m. Figure 1: Example of points provided from post inspection data processing Figure 2: Example Depth of Cover Report (Aerial imagery source: Esri, DigitalGlobe, GeoEye, Earthstar Geographisc, CNES/Airbus DS, USDA, USGS, AeroGRID, IGN and the GIS User Community)
46 PIPELINE TECHNOLOGY JOURNAL RESEARCH / DEVELOPMENT / TECHNOLOGY It can be seen that the OS Terrain data has a lower accuracy than the EA LiDAR data. The OS Terrain data has a lower stated accuracy for the product and also has a lower resolution. The consequence of lower resolution is that the detail of ground features is missing from the data, see Figure 5. This can be seen at distanc- es 2120 m and 2320 m. Figure 4: Accuracy Assessment of Ground Elevation Data ground elevation data To assess the accuracy of ground elevation data a comparison between infield measurements, EA LiDAR and OS Terrain 5 data was made. Figure 4 shows an accuracy assessment of the EA LiDAR and OS Terrain data against in field mea- surements. A ±0.07 m RMS error was calculated for the LiDAR data and ±0.46 m for the OS Terrain data. This is within the stated accuracy for each product. Ditch not evident in the OS Terrain Data Embankment not evident in the OS Terrain Data Figure 5: Comparison of Ground Elevation Data sets dePth of Cover To assess the accuracy of depth of cover results, a comparison between the estimated depth of cover and infield measurements was performed. Figure 6 shows the accuracy assessment using the pipe centerline data for all 10 pipe sections. This includes depth of cover calculated using EA LiDAR data and OS Terrain 5 data. The RMS error for depth of cover based on EA LiDAR data is ±0.15 m and for the OS Terrain data is ±0.46 m. Figure 6: Accuracy Assessment at all 10 Pipe Sections
RESEARCH / DEVELOPMENT / TECHNOLOGY PIPELINE TECHNOLOGY JOURNAL 47 CONCLUSIONS • • • The trial has successfully demonstrated ROSEN’s methodology to estimate the depth of cover over pipelines. This includes producing an accurate pipeline centreline from data obtained during a routine internal inspection, combined with ground elevation data available from the Environment Agency (EA) to calculate depth of cover. The results of the calculation have been validated against infield depth of cover measurements ob- tained using a pipe and cable locator. The accuracy of the depth of cover results has been calculated using a root mean square (RMS) error method. This has determined an overall accuracy of ±0.15 m us- ing EA LiDAR data. Infield ground surface measurements were com- pared with the EA LiDAR and OS Terrain data. A RMS error of ±0.07 m was calculated for the EA LiDAR data and ±0.46 m for the OS Terrain data. These show that the accuracy of the data is within the stated product specifications. References . UKOPA/GP/001. UK Onshore Pipeline Operators’ Association – Industry Good Practice Guide. Managing pipelines with reduced depth of cover. Edition 1, January 2016. . IGEM/TD/1 Edition 5. Steel Pipelines and Associated Installations for High Pressure Gas Transmis- sion. Institution of Gas Engineers and Managers, Communication 1735, 2008 . PD 8010-1:2015, Pipeline system – Part 1: Steel pipelines on land. Code of practice, British Standards, March 2015. . ASME B31.8-2014, Gas Transmission and Distribution Pipeline Systems, American Society of Mechanical Engineers, 2014. . AS 2885.1 – 2007, Australian Standard Pipelines – Gas and liquid petroleum Part 1: Design and construction. Standards Australia, 2007. . Environment Agency Geomatics Survey Data Website. Available at: https://environment.maps. arcgis.com/apps/MapJournal/index.html?appid=c6cef6cc642a48838d38e722ea8ccfee. Accessed 13th June 2017. . Ordnance Survey. OS Terrain 5 User Guidance and Technical Specification. Version 1.2 March 2017. Daniel Finley ROSEN Group Senior Engineer Authors Michiel Roeleveld ROSEN Group Data Analyst email@example.com firstname.lastname@example.org Simon Daniels ROSEN Group Principal Engineer email@example.com Paul Ogden National Grid Senior Engineer (Civil Assets) firstname.lastname@example.org Klaas Kole ROSEN Group Data Analyst email@example.com
48 PIPELINE TECHNOLOGY JOURNAL CONFERENCES / SEMINARS / EXHIBITIONS 14TH PIPELINE TECHNOLOGY CONFERENCE Pipeline T Europe’s Leading Pipeline Conference and Exhibition 19-21 MARCH 2019, ESTREL CONVENTION CENTER, BERLIN, GERMANY Conference 2010 EVENT PREVIEW 700+ DELEGATES 80+ EXHIBITORS 50+ DIFFERENT NATIONS From 19-21 March 2019 Europe’s leading conference and exhibition on pipeline systems, the Pipeline Technology Conference, will take place for the 14th time. ptc 2019 offers again opportunities for operators as well as technology and service providers to exchange latest onshore and offshore technologies and new developments supporting the energy strategies world-wide. More than 700 delegates and 80 exhibitors are expect- ed to participate in the 14th ptc in Berlin. The practical nature of ptc was always based on the cooperation with our technical and scientific supporters and on a top-class interna- tional advisory committee. The conference will feature lectures and presentations on all aspects surrounding oil, gas, water and product high, medium and low pressure pipeline systems. Please take a closer look into he “First Announcement and Call for Papers” and get involved now - send in your presentation suggestion and reserve your booth at the exhibition.
CONFERENCES / SEMINARS / EXHIBITIONS PIPELINE TECHNOLOGY JOURNAL 49 Pipeline T Conference 2010 14TH PIPELINE TECHNOLOGY CONFERENCE & EXHBITION EUROPE’S LEADING PIPELINE EVENT THE ANNUAL GATHERING OF THE INTERNATIONAL PIPELINE COMMUNITY IN THE HEART OF EUROPE After starting as a small side event of the huge HANNOVER MESSE trade show in 2006, the Pipe- line Technology Conference developed into Eu- rope’s largest pipeline conference and exhibition. Since 2012 the EITEP Institute organizes the ptc on its own and moved the event to Berlin in 2014. EXHIBITORS OF PTC 2018: 70+ Pipeline Operators 16 thematic focuses at ptc 2019 Construction Corrosion Protection Cyber Security Fiber Optic Sensing Inline Inspection Integrity Management Leak Detection Maintenance & Repair Materials Offshore Technologies Operational Improvements Planning & Design Pump & Compressor Stations Qualification & Recruitment Trenchless Technologies Valves & Fittings ptc Side Conferences on Qualification & Recruitment Public Perception ptc Seminars • Pipeline Life-Cycle Exten- • Geohazards in Pipeline sion Strategies • Inline Inspection • Offshore Engineering • Corrosion Protection • Pipeline Leak Detection AATS 2 5
50 PIPELINE TECHNOLOGY JOURNAL CONFERENCES / SEMINARS / EXHIBITIONS Platinum Sponsor Golden Sponsors Silver Sponsors Become a sponsor of the Pipeline Technology Conference and we will include your company in all our ptc marketing activities from the date of registration Pre-conference Marketing Adverts in Media Partner Journals ptc Website Pipeline Technology Journal (ptj) ptc + ptj Newsletter Press Releases Marketing at the Conference Conference Bag Brochure in Conference Bag Banners in the Exhibition Hall Banners in Conferences Rooms Booth in Preferred Location Post-Conference Marketing Press Releases ptc + ptj Newsletter Final Report Social Media Activities Brochures Letterhead Online Banner Direct Mailings Lanyard Event Smartphone App Get Together Sponsorship Dinner Sponsorship Social Media Activities Social Media Activities ptc Website Pipeline Technology Journal (ptj) EITEP Database of Verified Pipeline Addresses >50,000 EMAILS • • • highly international updated on a daily basis including all previous delegates, speakers, exhibitors Get in touch with us if you would like to get to know more about our sponsoring Opportunities.
CONFERENCES / SEMINARS / EXHIBITIONS PIPELINE TECHNOLOGY JOURNAL 51 CONFIRMED EXHIBITORS AS OF 03.09.2018 REGISTER YOUR STAND AT www.pipeline-conference.com/stand-booking 50+ DIFFERENT NATIONS DELEGATIONS FROM 70+ DIFFERENT PIPELINE OPERATORS FROM ALL AROUND THE WORLD 700+ DELEGATES 80+ EXHIBITORS 100+ PRESENTATIONS 25 TECHNICAL SESSIONS ACCOMPANYING SCIENTIFIC POSTER SHOW THEMATIC FOCUSES: CONSTRUCTION CORROSION PROTECTION CYBER SECURITY FIBER OPTIC SENSING INLINE INSPECTION INTEGRITY MANAGEMENT LEAK DETECTION MAINTENANCE & REPAIR MATERIALS OFFSHORE TECHNOLOGIES OPERATIONAL IMPROVEMENTS PLANNING & DESIGN PUMP & COMPRESSOR STATIONS QUALIFICATION & RECRUITMENT TRENCHLESS TECHNOLOGIES VALVES & FITTINGS
52 PIPELINE TECHNOLOGY JOURNAL CONFERENCES / SEMINARS / EXHIBITIONS JOB & CAREER MARKET YOUR OPPORTUNITY TO ATTRACT PROFESSIONALS AND HIGH POTENTIALS The international pipeline community is in need of additional personnel. We need more experienced pro- fessionals, but we also need young graduates to join our ranks. Despite attractive working conditions, many companies encounter problems while they are reaching out to potential re- cruits. There are many competing in- dustry sectors who are also in need of high potentials. This results in many vacant jobs in the pipeline community, for operators, technology providers and service providers alike. This necessity has driven us to develop a new service for the global pipeline indus- try. For this reason, we organize the first ptc side conference on Qualification and Recruitment. ptc side conference on Qualification and Recruitment 18 March 2019 Estrel Convention Center Berlin, Germany In the frame of
CONFERENCES / SEMINARS / EXHIBITIONS PIPELINE TECHNOLOGY JOURNAL 53 ONE SERVICE - MULTIPLE CHANNELS International Universities Offensive approach: We push forward and gen- erate attention to our career market directly at the universities. We also collect CVs from inter- national graduates and experts and forward it directly to you. Webseite Continuous promotion : Your vacancies are published on the Pipeline Technology Journal (ptj) website. In Addition, the ptj contains your vacancies too. Biweekly Newsletter Dead on target: We send your vacancies or your company profile to our database of 50,000 international pipeline professionals. International Events Physical appearance: The job & career market has an indi- vidual booth during all EITEP events. Questions? You get: Please contact Mr. Admir Celovic for further information and booking requests. firstname.lastname@example.org +49 / 511 / 90992-20 The most cost-effective support to your recruitment efforts available to the market
54 PIPELINE TECHNOLOGY JOURNAL COMPANY DIRECTORY Automation Siemens Germany www.siemens.com PHOENIX CONTACT Germany www.phoenixcontact.de/prozess Yokogawa Japan www.yokogawa.com Certification Bureau Veritas Germany www.bureauveritas.de Cleaning Reinhart Hydrocleaning Switzerland www.rhc-sa.ch/rhc/ Coating Denso Germany www.denso.de Kebulin-gesellschaft Kettler Germany www.kebu.de Polyguard Products United States www.polyguard.com Premier Coatings United Kingdom www.premiercoatings.com/ Shawcor United States www.shawcor.com TDC International Switzerland www.tdc-int.com TIB Chemicals Germany www.tib-chemicals.com Construction BIL Germany bil-leitungsauskunft.de Herrenknecht Germany www.herrenknecht.com IPLOCA - International Pipe Line & Offshore Contractors Association Switzerland www.iploca.com MAX STREICHER Germany www.streicher.de/en Petro IT Ireland www.petroit.com VACUWORX Netherlands www.vacuworx.com Vlentec The Netherlands www.vlentec.com Construction Machinery Maats Netherlands www.maats.com Worldwide Group Germany www.worldwidemachinery.com Corrosion Protection TPA KKS Austria www.tpa-kks.at Engineering ILF Consulting Engineers Germany www.ilf.com KÖTTER Consulting Engineers Germany www.koetter-consulting.com
PIPELINE TECHNOLOGY JOURNAL 55 COMPANY DIRECTORY Inline Inspection Integrity Management 3P Pipeline, Petroleum & Precision Services Germany www.3p-services.com A.Hak Industrial Services Netherlands www.a-hak-is.com KTN AS Norway www.ktn.no LIN SCAN United Arab Emirates www.linscaninspection.com NDT Global Germany www.ndt-global.com Pipesurvey International Netherlands www.pipesurveyinternational.com PPSA - Pigging Products and Services Association United Kingdom www.ppsa-online.com Romstar Malaysia www.romstargroup.com Rosen Switzerland www.rosen-group.com Inspection Ametek – Division Creaform Germany www.creaform3d.com Applus RTD Germany www.applusrtd.com EMPIT Germany www.empit.com Metegrity Canada www.metegrity.com Pipeline Innovations United Kingdom www.pipeline-innovations.com Leak Detection Asel-Tech Brazil www.asel-tech.com Atmos International United Kingdom www.atmosi.com Direct-C Canada www.direct-c.ca Entegra United States www.entegrasolutions.com GOTTSBERG Leak Detection Germany www.leak-detection.de MSA Germany www.MSAsafety.com/detection OptaSense United Kingdom www.optasense.com Pergam Suisse Switzerland www.pergam-suisse.ch PSI Software Germany www.psioilandgas.com sebaKMT Germany www.sebakmt.com SolAres (Solgeo / Aresys) Italy www.solaresweb.com
56 PIPELINE TECHNOLOGY JOURNAL COMPANY DIRECTORY Materials egeplast international Germany www.egeplast.de Monitoring Krohne Messtechnik Germany www.krohne.com Pump and Compressor Stations TNO The Netherlands www.pulsim.tno.nl Repair CITADEL TECHNOLOGIES United States www.cittech.com Clock Spring United States www.clockspring.com RAM-100 United States www.ram100intl.com T.D. Williamson United States www.tdwilliamson.com Research & Development Pipeline Transport Institute (PTI LLC) Russia www.en.niitn.transneft.ru Further boost your brand awareness and list your company within the ptj - Company Directory www.pipeline-journal.net/advertise Safety DEHN & SÖHNE Germany www.dehn-international.com/en HIMA Germany www.hima.de TÜV SÜD Indutrie Service Germany www.tuev-sued.de/is Standards & Regulations DNV GL Norway www.dnvgl.com DVGW - German Technical and Scientific Association for Gas and Water Germany www.dvgw.de Surface Preparation MONTI - Werkzeuge GmbH Germany www.monti.de Trenchless Technologies GSTT - German Society for Trenchless Technology Germany www.gstt.de Rädlinger Primus Line Germany www.primusline.com Valves & Fittings AUMA Germany www.auma.com IMI Precision Engineering Germany www.imi-precision.com Zwick Armaturen Germany www.zwick-armaturen.de
14TH PIPELINE TECHNOLOGY CONFERENCE Europe’s Leading Pipeline Conference and Exhibition 19-21 MARCH 2019, ESTREL CONVENTION CENTER, BERLIN, GERMANY www.pipeline-conference.com Next Issue: October 2018 Pipeline Technology Journal In the next Edition of ptj: Offshore Technologies www.pipeline-journal.net Event Calendar International Pipeline Conference 2018 24 - 28 September 2018 Calgary, Canada International Pipeline Expo 2018 25 - 27 September 2018 Calgary, Canada gat | wat 2018 23 - 25 October 2018 Berlin, Germany ptc Side Conference on Qualification & Recruitment ptc Side Conference on Public Perception 18 March 2019 Berlin, Germany 18 March 2019 Berlin, Germany 14th Pipeline Technology Conference (ptc) 19 - 21 March 2019 Berlin, Germany
Pipeline Technology Journal You have interesting content to share with the global pipeline community? You want to enhance or maintain your international visibility as a company? Submit an Article! Book an Advertisement ! Use ptj as a platform to report about your news, projects, innova- tions and technologies. If you are interested in submitting insightful technical articles to be considered for the ptj, please send us an abstract for review. North America 37,8% Europe 33.4% Mena Region 6.8% Africa 2.5% South America 4.5% Asia 12.6% China special e-mail list of 20.000 recipients Oceania 2.5% The ptj-brand offers a multitude of advertise- ment opportunities to increase visibility and reputation toward pipeline professionals worldwide. Make use of the extensive ptj-portfolio and reach over 50,000 Experts. ptc ADVISORY COMMITTEE / ptj EDITORIAL BOARD ptj-brand-audiences CHAIRMEN Heinz Watzka, Senior Advisor, EITEP Institute Dirk Strack, Technical Director, TAL - Deutsche Transalpine Oelleitung MEMBERS Ulrich Adriany, Senior Technical Expert, ARCADIS Deutschland Arthur Braga, Country Manager, ITF Brazil Jens Focke, CEO, BIL Maximilian Hofmann, Managing Director, MAX STREICHER Dirk Jedziny, Vice President - Head of Cluster Ruhr North, Evonik Industries Dr. Andreas Liessem, Managing Director, Europipe Ralf Middelhauve, Head of Central Dept. Process Industrie / Plant Engineering and Operation, TÜV NORD Systems Dr. Prodromos Psarropoulos, Structural & Geotechnical Engineer, National Technical University of Athens Uwe Ringel, Managing Director, ONTRAS Gastransport Muhammad Sultan Al-Qahtani, General Manager, Pipelines, Saudi Aramco Filippo Cinelli, Senior Marketing Manager, Baker Hughes, a GE company Andreas Haskamp, Pipeline Joint Venture Management, BP Europa SE Dr. Thomas Hüwener, Managing Director Technical Services, Open Grid Europe Cliff Johnson, President, PRCI - Pipeline Research Council International Michael Lubberger, Senior Prod- uct Manager Pipeline, BU Utility Tunnelling, Herrenknecht Steffen Paeper, Offshore Engineering, South Stream Juan Arzuaga, Executive Secretary, IPLOCA Dr. Marion Erdelen-Peppler, Secretary General, EPRG - European Pipeline Research Group Jörg Himmerich, Managing Di- rector / Technical Expert, Dr.-Ing. Veenker Ing.-ges. Mark David Iden, Director, Gov- ernment Relations, Floating Leaf Mike Liepe, Head Business Solu- tion Line O&G Pipelines, Siemens Brigham McCown, Chairman and CEO, Nouveau Bruno Pomaré, Technical Director, Spiecapag Frank Rathlev, Manager of Network Operations, Thyssengas Dr. Joachim Rau, Managing Director, DVGW CERT Hermann Rosen, President, ROSEN Group Michael Schad, Head of Sales International, DENSO Audience Job Levels 11% CEO 20% Director 26% Manager 43% Executive Company types 23% Operators 61% Techn. / Service Providers 11% Researchers 4% Authorities Dr. Adrian Schaffranietz, Coordi- nator Government Relations, Nord Stream 2 Prof. Dr. Jürgen Schmidt, Manag- ing Director, CSE Center of Safety Excellence Ulrich Schneider, Business Development Manager Continental Europe, KTN Guntram Schnotz, Expert / Pipeline, TÜV SÜD Industrie Service Carlo Maria Spinelli, Technology Planner, eni gas & power Anand Kumar Tewari, Executive Director, Indian Oil Corporation A manifold database Asle Venas, Senior Pipeline Special- ist, DNV GL Paul Waanders, Int. Sales Manager, Maats Pipeline Equipment George Ziborov, Leading expert, Foreign Economic Relations Department, Transneft Bernd Vogel, Head of Network Department, GASCADE Gastransport Tobias Walk, Director of Projects – Pipeline Systems, ILF Consulting Engineers Roger Vogel, Sales Manager - EURA, Baker Hughes, a GE company Thomas Wolf, CEO, NDT Global We deliver content to local practicioners and global decision-makers alike, making the ptj- brand a suitable tool for global knowledge dis- tribution as well as developing and upholding overall visibility in the global pipeline industry.