Abstract Leaks and ruptures are the most important possible risks for operational oil and gas pipelines. Due to their hazardous effects on the environment, much research has been conducted to prevent and detect possible ruptures on a pipeline to protect people and the environment by enhancing safe operation. Any improvement in leak detection technologies to increase the accuracy and sensitivity while eliminating false leak and rupture alerts will protect the environment and assure hazard-free operation. Data mining algorithms are widely used in many industries, including the energy industry. They have already been implemented as computational leak detection methodologies. To increase confidence and improve accuracy and sensitivity, different algorithms may be introduced to detect ruptures. In our study, a 36" crude oil pipeline with two pump stations was configured in a pipeline simulator. The pipeline parameters of flow, pressure, and temperature were computed for several leak and rupture cases, and data science algorithms such as Logistic Regression, Neural Network, and Multivariate Adaptive Regression Splines were used as classifiers to detect the leaks and ruptures. Multivariate Adaptive Regression Splines (MARS) is an important statistical learning tool for both classification and regression. MARS is nonparametric, adaptive, and effective in high dimensional problems with a proven record for fitting nonlinear multivariate functions. The contribution from the basis functions together with interaction effects between the predictors are used to determine the response variable: MARS produces a resultant model as an explicit formula. MARS proves itself as a comparative classifier to the already known logistic regression and neural network methods as a new leak and rupture detection computation data science technique for pipeline operators.

Abstract In September 2018, a field engineer’s mistake while replacing cast iron main in Lawrence, Massachusetts (USA), led to a death and over a billion dollars of property damage. Residents and businesses in the area were without natural gas service for months, at a time of the year when temperatures had already started dropping. One year later, every natural gas utility in the United States is bearing a share of the brunt of the backlash, as the National Transportation Safety Board starts to lay the groundwork for increased regulations nationwide. One aspect of the heightened attention on natural gas utilities is the vulnerability of the regulator stations themselves, even though the cause of the disaster cited above was traced back to a misplaced sensing line. Regulator stations are such a critical element of a distribution network that any comprehensive risk analysis program must include them. One tool that could help is a regulator station risk model that evaluates the potential fail open and fail closed scenarios for numerous regulator station design types. The likelihood of these events is then paired with the potential consequences, including health and safety, property damage, and network outage impacts. The output of these risk models form a critical set of inputs for use in an overarching probabilistic risk model, as being promoted by the Pipeline and Hazardous Materials Safety Administration (PHMSA), to support risk-informed decision-making. The early part of this paper describes the regulator station risk model itself. The latter part documents the methodology used to provide critical information required by the risk model for determining consequence of failure. Finally, a parametric study approach is added to analyze the hydraulic consequences of regulator stations completely failing open. This paper discusses how the application generates results, including number of customers affected, how much pressure is seen downstream, station scoring, and total length of pipe overpressured. This analysis is performed wholly on Microsoft Azure public cloud, using a version of a commonly-used hydraulic engine from a desktop product. This architecture enables an entire system to be analyzed in minutes compared to hours or days for an engineer to complete manually. Time is always a barrier to performing lengthy analyses, and this application helps to erode that barrier by enabling utilities to generate the results they need to perform risk analyses more frequently. This is important because these natural gas systems are constantly in states of disrepair or repair. Having an up-to-date risk model helps to keep the public safe and helps the utility to make the best decisions for managing and operating their system. DNV works directly with clients to implement necessary measures to define threats and comply with federal and local regulations. The paper’s authors intend on partnering with select utilities to develop comprehensive station risk management plans, which will help refine the methodologies described in the paper. This may not take place until after this paper is published, so any advancements derived from these partnerships may have to wait for publication at the 2021 conference.

Abstract A pressure surge in a pipeline occurs following a sudden valve closure, compressor start-up or trip. Pipeline simulations can be used to model the physical surge as it travels down the line. As it travels, the amplitude of the pressure wave decays, and the slope of its leading edge may decrease due to friction or to effects related to compressibility. These surges can be problematic for automatic emergency shutdown (ESD) valves. ESD valves are configured to close automatically during a pipeline rupture to minimize spillage. These valves can also trigger on the pressure surge from a leak. ESD valves must be calibrated to trigger from a rupture or large leak, but not from the surges that arise during normal operations. Determining how to do this presents a challenging problem for pipeline simulation. The simulator must give an accurate picture of the surge, but the true physical dispersion of the surge due to friction, if any, can easily be saturated by the numerical dispersion inherent in the pipeline model. This paper investigates the numerical dispersion of pressure waves in a gas pipeline, and what can be done to minimize that dispersion. We examine both small-amplitude surges such as those coming from compressor trips and large-amplitude surges such as those that might be generated from a pipe rupture. In order to distinguish the true physical behavior of pressure surges from numerical effects, we compare the results of several types of simulator. The Preissman box scheme [1], a widely used implicit method which was first developed for simulating open-channel flow, will be compared with the MacCormack predictor-corrector method [2] and a higher-order extension of the box scheme. The advantages and deficiencies of each of the methods will be discussed, as well as ways to achieve accurate simulation of this sort of surge with simulation methods commonly found in commercial pipeline simulators.

Abstract This paper investigates the technology and business benefits of a web-based model scheduling application, which integrates a transient model to other business applications that provide injection and demand schedules. The application provides a simple method to input complex scenarios, including compressor, regulator, and routing operations. The paper shows how the system allows non-modellers to use a complex transient model, reducing non-optimal operational decisions which could impact the wholesale natural gas price or supply to end user customers such as households or businesses. Other operational metrics are discussed. Introduction and Background Independent System Operators’ (ISOs) and Transmission System Operators’ (TSOs) use of transient pipeline models to manage nominations and allocations is seemingly an easy operational premise. The Australian Energy Market Operator (AEMO) operates the gas Declared Transmission System (DTS) which serves Victoria, a state in Australia. The DTS network consists of approximately 2,200 km of transmission pipelines with various lengths and diameters, and transports gas to a network from up to 10 different suppliers. The high pressure meshed transmission network supplies gas to 148 custody transfer points which spans the geographical extremities of the state. The DTS is becoming increasingly more complex, with changes to flow paths and dynamic demand regimes. AEMO, like other TSOs and ISOs, is utilising transient pipeline models to inform operational strategies. This paper explores the challenges facing ISOs and TSOs in the United Kingdom, The United States, and Victoria Australia. Specific reference is made to AEMO’s gas transmission network. The issues explored cover a range of subjects including financial and macro policy but all lead to the impact on a transmission network. The paper does not attempt to solve each transmission operator’s challenges in detail. However, within the context of AEMO’s operations, it attempts to identify the similarities with other TSOs and highlight how AEMO’s transient hydraulic model helps it deal with its challenges.

Abstract The integration of Renewable Energy Sources (RES) into the electric power system and the demand for efficient, clean and reliable production, transport and distribution of energy is connected with an increased coupling between the individual energy systems. The interdependency between natural gas and electric power systems, for instance, is increasing and expected to increase further in the near future, mainly due to the growing demand for flexible backup generation and the need for short- and long-term energy storage options to balance the fluctuations of variable and intermittent RES. The need for flexible backup power can be covered by Gas Fired Power Plants, while the demand for seasonal storage can be met by Power-To-Gas facilities. In this paper, we show how curtailment of RES can be reduced by operating Power-To-Gas facilities. Introduction To meet future decarbonization targets power system planners and operators have to integrate more variable renewable energy sources into their electric power systems. The stable and reliable operation of the power system, however, requires a balance between power generation and demand, since electric energy cannot be efficiently and economically stored in existing power system networks. Due to the uncertainty and variability in the generation from Renewable Energy Sources (RES) and at the same time the lack of energy storage, power system planners and operators must find new flexibility and storage options, which typically lie at the interface between different energy networks, such as natural gas, heat & power networks. Natural gas networks, for instance, have relatively large energy storage capacities (linepack, underground gas storage, LNG storage tanks), which can be leveraged by electricity networks to tackle the flexibility and energy storage challenges. The connection between the gas and electric power system can exist at Gas-To-Power plants (G2P) as Gas fired Power Plants (GPP), electric driven gas compressor stations (Power-To-Pressure - P2P) as well as Power-To-Gas facilities (P2G). G2P facilities have relatively short start-up and shut-down times, making them well suited as back-up generation sources in case of lack of wind and solar generation. P2G facilities, on the other hand, can be used to convert excess power generation from renewable sources into hydrogen (H2) and/or synthetic natural gas (SNG), which in turn can be injected into the natural gas pipeline network and/or underground gas storage reservoirs.

ABSTRACT Heavy oil accounts for more than half the word's total oil reserves. However, heavy oil as extracted from the reservoir has an extremely high viscosity and the transportation of these fluids in a pipeline requires the use of technologies to lower the viscosities of this oil. Among these technologies are heating or dilution of the heavy oil and sometimes a mixture of both of these methods. The majority of oil storage terminals contain pipes that are installed above ground and in northern countries are prone to experiencing cold temperatures during winter months. In part of the terminal where the fluid stays stagnant in the pipes (such as relief piping and manifold pipes) the cold weather can impact the viscosity of the fluid, even after it has been diluted. When the pumping system has to move fluid that has been stagnant in the pipe and has become colder, the pressure required to achieve this is very high. The transient surge pressures created in these cases can become extreme to the point of exceeding the maximum operating pressures of the pipe. When performing transient simulations, it is important to consider the fluid with higher viscosities that have been kept in the pipes and that will require higher pressures to pump them down the line, so that pressure surges can be more accurately calculated and mitigated. This paper will present comparisons of how the transient pressures change between cases of heavy oils at different temperatures and viscosities. It will show that if heavy oil is kept stagnant in a pipe that is subject to cold temperatures, the pressures required to push this fluid down the pipe increases as the viscosity becomes higher. IN particular, it will present the case where the relief piping has been installed above ground and when the fluid needs to be displaced, the surge pressures can become significantly higher than expected.

ABSTRACT LNG Terminals represent key infrastructure for gas markets in many regions. With markets shifting to a more global approach, what happens when your operational cases evolve, and the original design specifications are challenged? Or worse an operational event exposes weakness in the design? The priorities on any site continue to be around safe, profitable operation and maximising capacity/uptime. Under normal operation piping systems can withstand the resulting stresses. Events such as rapid PERC (powered emergency release coupling) closures, Emergency Shut Downs (ESDs), valve operations, and pump trips can all create transient effects. Consequences can include excessive surge pressures which in turn lead to increased pipe stresses and structural loads. An ongoing relationship with a large LNG terminal operator has driven studies into hydraulic surge in pipelines following Emergency Shut Down (ESD) events and developed this hydraulic understanding to include bend forces and resolved pipe forces which could be passed into targeted structural analysis. Over time, scenarios have become more complex including for example an increasing need to be able to re-load ships from a site originally designed as an import terminal. Alongside the technical challenge of these studies; a key factor was generating justifiable rationalisation of all permutations of operating scenarios into a manageable quantity of worst-case simulations for a multi-tank and multi-jetty site. Combined with the increasing complexity of cases we were required to develop methods to run simultaneous simulations and manage the model files and outputs in a reliable fashion. More recent work has led to feeding the force outputs into static and dynamic load analysis to understand the force and harmonic interactions with the physical structures on the site. Pressure and force results were extracted for each scenario using a high-fidelity hydraulic model and then utilized in structural analysis tools for further analysis. This paper outlines the evolution of a complex modelling problem, and the effects of working on an inherited model from a 3rd party and architecture built on scripting tools to solve problems now integrated into our modelling solution. With the concept of simultaneous modelling runs and parametric studies commonplace in today's environments, we consider the progression of this approach to meet future requirements.

ABSTRACT A novel model is used to predict the inception moment and location of a phenomenon known as the column separation or slack line. The existing models for prediction of the cavitation phenomenon in internal flow systems in general are primarily formulated for the capture of the inception moment and initial location of the event only. Hence questions such as where the incepted cavities and bubbles tend to go, or the detailed state of a slack line in time should be addressed using a different approach. The model presented in this paper is validated and verified against experimental data available in the literature before being applied to a 53 Km (33 miles) industrial pipeline. Imposing conditions at the injection and delivery ends of the line tests various scenarios causing column separation. Hydraulic parameters such as pressure head and flowrate as well as interfacial mass transfer rate of the incepting and collapsing bubbles and cavity zones are predicted in real time over the entire domain of space and time. The results predicted the fate of the separated column of the fluid, whether they rejoin or continue to change the size with different rates or even if they become stabilized stationary cavity pockets after passage of minutes or hours. Results are also compared based on the initial cause of the column separation such as huge transients. Also, accurate prediction of the whole column separation event, distinguishes it from other events, which would commonly mimic or mask the column separation by exhibiting identical hydraulic footprints.

ABSTRACT As the gas industry landscape gets ever more complicated with thousands of new players entering the gas business, the challenges for efficient management of day-to-day operations are mounting. Situation on the gas market with ever changing nominations for transport pose significant problems for TSO's to change their network mode (compressor stations, regulators, valves) to match the today and tomorrow's demand. There is a need to dynamically manage network switching and network elements to match the situation when nominations can dramatically change every hour. Gas companies are looking for a tool that can optimise the current situation and nearest future with as low as possible transport cost while fulfilling the contract constraints. We have found that after talking to a number of TSO's that the actual question posed by dynamic optimisation varies from one company to another. Therefore, it is impossible to create one universal algorithm that would fulfill the needs of everyone (apart from brute force, but that is not a task for today's computers). We have devised a way how to supplement this lack of raw computing power with human interaction and knowledge of individual network. We will talk about our experience of creating a tool that assists control room dispatchers in quick reaction to changing transport situation. This tool we call Net Schedule Management (or NSM in short) and it consists of two modules - one that archives a gas day simulation and cleans bad SCADA data from it and one that actually optimises the network mode. To dynamically optimise network, operator can pull a historical day that most matches the current situation and predicted consumptions and transport, simulate different scenarios and try to improve the criteria of fuel gas while observing the pressure limits. Steady state optimisation can help him as well as a library of network switching, but ultimately it is the human experience and knowledge of the network that does the trick. We will present a case study that confirms 4% - 23% daily savings of fuel gas on a medium complexity gas network using the NSM tool.

ABSTRACT Upstream gas supply planning and allocation contains 3 cycles namely long term (20 years - end of life of a field) mid-term (up to 1 year) and short term (72 hours). Since the upstream assets are owned/operated by different companies under various production sharing contracts, each of the operating companies prefer to maximize their production and recover the investment costs at the earliest. A regulator acts as a custodian to ensure that all stakeholder's entitlements are protected. The custodian relies on economic and hydraulic modelling to analyse the production plateau's shared by various operating companies together with the demand profiles for viability. The commercial planning team of the custodian works closely with flow assurance engineers before confirming the production numbers as the planners are unaware of any capacity constraints in the physical network for flow. In the short term, an integrated network operator breaks the long term demand numbers and controls the network for seamless delivery of gas from upstream to LNG plants (mostly). The purpose of the paper is to highlight the challenges faced by planners and engineers during the planning cycles and to make appropriate recommendations.

ABSTRACT New designs or modifications of existing pump or gathering facilities are required frequently in many industries due to new operating conditions, increase in product demand, pump-pipeline interaction issues, pulsations problems, Net Positive Suction Head (NPSH) requirements, and so forth. In many cases, harsh transient events are experienced due to the lack of proper design, change in operating conditions, or unexpected process conditions. Piping systems are frequently subject to transient events such as slugs, water hammer, and cavitation that can create high amplitude forces and pressure spikes. Liquid systems are more susceptible to damaging forces during transients than gas systems due to the high density and incompressibility of the operating fluid. These events often result in high impact forces and vibrations that sometimes cause failures. This paper presents a general approach for modeling critical transient events in liquid and multiphase piping systems. In addition, case studies will be presented to identify the critical areas of the modeling, show model-field data comparison, results interpretation, and present some possible mitigation actions. INTRODUCTION AND BACKGROUND Transient events can introduce large pressure forces and rapid fluid accelerations into large pipeline systems, small manifolds, or distribution systems. These disturbances may result in different types of failures in pump, pipes, and devices such as pipe rupture or components reduced life due to cycle fatigue. Many transient events can lead to column separation or water hammer, which can result in catastrophic pipeline failures. Thus, transient flow simulation has become an essential requirement for increasing reliability and ensuring the safe operation of different pipeline systems and processes. Liquid systems are sometimes difficult to analyze for transient events as these can be dependent on factors that are not always known or finalized in the design stage or change during the lifetime of the system. Dependent variables include the characteristics of the fluid, the piping geometry, pressure reducing valves, operating conditions and nature of the events, scenarios and flow rates. Various modeling tools exist to provide reliable predictions and cost-effective solutions for hydraulic transient events, vibration or pulsation problems, but often multiple types of analyses must be used in combination to evaluate the system as a whole or solve complex problems. The traditional method of analyzing transient events in liquid systems is to only use commercially available software to predict the extreme pressure conditions; however, often these transient fluid models need to be combined with mechanical or thermal stress analyses; dynamic pressure versus time inputs, and sensitivity studies or field data evaluations to be effective in resolving problems and ensuring pipeline integrity.

ABSTRACT The objective of this paper is to describe a method that simulates an energy recovery system (ERS), which exploits water hydraulic power to boost inlet flow pressure. The impact of pipeline pressure surge (water hammer) on water treatment units was investigated. Surge pressure and pressure rise rate were calculated. A novel methodology has been developed in this paper to simulate an energy recovery system and estimate pressure rise rate. This method integrated an energy recovery system into an existing pipeline simulation model. The energy recovery system model was developed using basic hydraulic pump equations. Actual system efficiency was used. Both maximum surge pressure and pressure rise rate are calculated each model time step. This same method can be used for other energy recovery systems hydraulic analysis. In this study, a high-pressure feed pump with a discharge pressure of 630 psig was analyzed. The model was used to calculate the maximum surge pressure downstream of the ERS. In this analysis, downstream of the ERS there is an RO (reverse osmosis) filtration system. The maximum pressure and rate of change of pressure must be controlled so as not to damage the filter membranes. Different surge scenarios were investigated. For the cases analyzed it was possible to keep the maximum surge pressure to 1117 psig that is below the maximum membrane design pressure. It was also possible to keep the maximum pressure rise rate for all cases simulated to below 5.2 psi/second. The membrane warranty for the cases analyzed limited the pressure rise rate to 10 psi/second and stipulated a maximum pressure or 1200 psig. The simulation results also provide design parameters for sizing surge relief devices and designing the required control system. Traditional surge analysis tools can properly estimate surge pressure within the pipeline system. However, energy recovery system behavior in a surge scenario was not simulated previously. The provided method can simulate energy recovery systems, calculate maximum surge pressure and pressure rise rate. The method sheds light on simulating energy recovery system and can be adopted for different simulation tools.

ABSTRACT When designing a piping system, normally the inherent control valve characteristics, e.g. linear or equal percentage valve opening/closing curves, are considered. However, inherent valve curves only consider the control valve as a "bobble". The characteristics of the valve will change once it is installed with piping connections, meters, equipment, or other valves and fittings. The additional friction loss introduced by piping connections or valve combinations is normally a function of the flow rate instead of staying as constant. This will change the overall opening and closing characteristics of the control valve. It is well known that surge pressure is directly related to valve characteristics. The combination of control valve with other components may create undesirable surge scenarios in operation which is commonly neglected in the design. This paper examines how the connections of the control valve with other piping components can influence the installed valve characteristics and surge pressure level in valve closings. The focus is on two aspects: how other components such as an ESD valve immediately upstream or downstream can influence the surge behavior of the control valve closing; how the upstream or downstream control valve influences the surge behavior of the ESD or Mainline Block Valve closing. The paper will present how the installed valve characteristics are different from the inherent characteristics and how significant the increase in the pressure surge was developed. The results and conclusions provided in this paper will serve as a general guideline for valve arrangement and piping design for reducing potential surge pressure in liquid systems. INTRODUCTION AND BACKGROUND In piping design the control valves present unique influence to system hydraulics resistance. It is well known that once installed in the piping system, the control valve characteristics (the relationship between valve flow coefficient and valve opening will change) (Sines, 2009, Headley 2003). A so called installed valve coefficient is introduced to describe this behavior. The valve coefficient is normally tested in the shop as a "bobble".

Abstract The depressurization of a pipeline containing a subcooled liquid at high pressure is an operation to be considered and analysed during pipeline design, in order to verify the relevant time trends of the hydraulic parameters, with particular focus on temperature. Usually, considering onshore pipeline systems, the blowdown operation is carried out for a section included between two MLV stations only, as a consequence of an emergency or a damage repair scenario; the inventory is discharged through vents or flares installed at the adjacent MLV stations. The types of fluids considered in this study are classified as HVP (High Vapour Pressure fluids), and include LPG, NGL, ethane rich mixtures, CO 2 rich mixtures. In this study, pipeline depressuring behaviour of HVP fluids is investigated by performing a sensitivity on the main governing parameters as fluid thermodynamic properties, heat transfer from environment, pipe wall thickness, pipeline elevation profile, as well as on the influence of discharge rates and discharge location configuration (one side, two-side). The analysis is performed by using the commercial hydraulic simulation tool OLGA, with full consideration of possible two-phase flow regimes, liquid holdup distribution, heat transfer modelling including the transient soil thermal dynamics. The target of this study is to provide indications for the engineering analysis of such kind of pipelines. Introduction The analysis of the time evolution of an industrial scale onshore buried pipeline containing a HVP fluid during a potential depressurization operation, performed by using flaring or cold venting facilities with realistic discharge characteristics, and taking into consideration the effects of hilly terrains, pipe wall thickness and heat transfer from soil cannot be found in literature. On the other hand, published papers present simulation of short and small size pipe sections (length of 100 m, 328 ft) typical of testing facilities, compared with the reference blowdown tests of Isle of Grain [1], [2], where the duration of the fluid discharge experiments were few tens of seconds, or idealized long pipelines with completely flat profiles, homogenous two-phase flow and external heat transfer coefficients typical of offshore pipelines or adiabatic scenarios, subject to full-bore ruptures [2], [3]. In those models important factors influencing real onshore pipeline systems depressurization evolutions as long liquid residence time in dips, long time duration of the operation (can be order of some days), thermal heat capacity of the steel and soil thermal impact are not captured. For HVP liquids, the physical behaviour during blowdown of pipeline sections of typical lengths of the order 20-40 km (12.4-24.8 miles) is similar. In fact, all exhibit a quite rapid initial depressurization following a thermodynamic expansion within the liquid phase region; then the fluid crosses the bubble point curve and flashes, and for a significant time the evolution is close to isothermal and isobaric, since the fluid behaves in a way very close to a one-component substance. During this period characterized by long pressure and average temperature plateaux, the fluid progressively vaporizes and tends to achieve the full vapour phase condition, while flowing as a two-phase mixture.

Abstract In recent years, the electric power industry has benefited from decreases in natural gas prices. These low prices have encouraged the installation of natural gas-fired power plants and have increased connections between natural gas and electric power systems. However, despite their dependencies, the gas and power industries have technical, political and commercial constraints that often force them to plan, operate and manage in isolation. Such limitations sometimes lead to undesirable outcomes, such as those experienced by both systems during the winter of 2013/2014 in the northeastern United States. In this paper, we focus on reducing the technical and computation limitations. We first present a mathematical model of combined gas and power systems and apply it to planning problems where expansion and upgrades are needed to meet increased demand. We second describe our previous work in convex relaxations that have yielded promising results. Finally, we introduce a gas-price elasticity model that includes the effects of gas-price volatility caused by congestion of the integrated gas-grid expansion system. Introduction Recent advances of cheap natural gas extraction technologies have led to record low prices for natural gas during the last decade. This drop in price has led to shifts from other fuels used by electric power systems to natural gas to produce electricity. In most cases, the fuels used by generators in power systems are procured days in advance. More importantly, perhaps, there is limited competition for the fuel. However, procurement of natural gas is different. Often, gas-fired power generators do not have firm natural gas contracts and procure their gas on the open market a day in advance at most. Thus, they are exposed to significant competition from other sectors (such as heating). As a result, while cheap natural gas has driven the price of electricity down, it has also increased fuel supply risk in power systems. Recent experiences in the northeast United States, such as during the winter of 2013/14, have highlighted this risk. Record cold temperatures directed gas supply to heating applications and left power systems scrambling for alternate sources of power. In this case, there were no blackouts due to generation shortages, but there were significant economic consequences [1], [2]. The risk associated with increased reliance on natural gas is influenced by a number of factors, of which one is the separation in management and operation of gas and power systems. This separation exists for a number of reasons. From a technological perspective, problems that join both systems are computationally very difficult and discourage joint management. In addition, in many places, the sectors are prohibited from coordinating activities or disjointed market structures discourage coordination.

Abstract The recent increase in the use of natural gas as a fuel for electricity production has greatly strengthened the interdependence between the electric power and natural gas industries. The resulting intra-day fluctuations in pipeline loads caused by intermittent or unexpected gas-fired electric power plant operations have become particularly problematic. To mitigate this issue, regulatory changes have been initiated to lower the barriers on communication between operators of power grids and gas pipelines. In this paper, we examine some types of intra-day operational information that pipeline managers can obtain from power grid operators in order to forecast time-varying pipeline loads and characterize uncertainty of these loads in space and time. We also present methods that could be developed by the pipeline simulation industry to more effectively use such information to mitigate intra-day gas-electricity interdependence issues. Specifically, we describe new techniques for using transient optimization to find feasible compressor operation schedules, which are resilient under uncertainty, given such power system data. Bringing these methods into practice can enable gas pipelines to reduce the risk of service interruptions caused by intermittent power plant activity. Additionally, pipelines will be able to more reliably service power plants that use their gas nominations during only part of the day, or that may start up or shut down with little warning. Introduction Extensive gas-fired power plant construction, as well as subsequent utilization of these plants to serve peak electric loads and provide generating reserves, has caused electric power grids to increasingly depend on greater and more reliable gas supplies. Such gas-fired generation can be intermittent, which causes large and sudden variations in takes from high pressure gas pipelines. These conditions lead to gas price fluctuations, line pressure drops, operational flow orders, and increased operating expenses for both industry sectors. Because natural gas is now the largest fuel source used for electric power production in many regions of North America, interest in coordinating the operations of these systems has grown. However, natural gas pipelines and power grids operate on very different spatial and temporal scales, and this makes the coordination of market clearing procedures and physical infrastructure operations difficult. While gas is purchased using nominations for steady takes over 12 to 24 hour intervals, power system operators often require gas-fired generators to commit to production schedules in which their nomination is burned over only part of the contract period. Furthermore, this schedule may change unexpectedly because of real-time re-dispatch and reserve generator activation. Thus, pipelines must use line pack and storage to balance supply rates with variable and uncertain delivery volumes. In light of this, pipeline system managers would benefit greatly from information about upcoming or possible changes in gas takes. Adapting pipeline operations to maximize efficiency and security under these new conditions requires simulation and optimization methods that accurately account for transient flows. It is well-understood that transient optimization methods for managing time-varying flows throughout intercontinental pipeline systems require a predictive element to be effective, because changes in one pipeline zone may take many hours to be felt in a different, distant zone. Specifically, information about the volume, timing, and uncertainty of variable and intermittent gas takes is necessary. These variations are currently caused primarily by the commitment and re-dispatch of gas-fired generators.

Abstract This paper will discuss and demonstrate technology developed by PG&E and GTS to address the unprecedented increase in safety work occurring at PG&E. The new technology, called Batch Analysis Tool (BAT) enables the planning engineer to analyze systems under a multitude of demand and operating conditions as a "batch" of simulations. User-defined scenarios are performed and cataloged in an automated, sequential manner rather than manually, one run at a time. This approach allows for the performance of hundreds or thousands of scenarios accurately and rapidly. This process also aids in optimizing operations and uncovering non-intuitive solutions. The benefit of this technology extends beyond companies with large amounts of safety work to include any study that requires a range of inputs to be hydraulically analyzed. The significant benefits Business justification and strategic technology are the primary emphases of this presentation. Representative use cases of the new technology are presented by PG&E in PSIG papers 1614 and 1615. Introduction and Background A planning engineer's job is to perform hydraulic analysis studies of pipeline networks. Hydraulic studies rarely only require 1 or 2 simulation runs to provide a complete study. Instead, a range of runs or scenarios must be performed to provide a sufficiently thorough study. Because of this, planning engineers run many simulations over the course of month, year, or career. Given the large number of runs performed significant benefits can be gained my making the task of performing simulation runs more time efficient. This can not only result in labor savings but also allow for more runs to be performed, increasing the thoroughness and accuracy of the study. The pace of safety work on PG&E's gas system has greatly increased the amount of hydraulic analysis and simulation runs required by planning engineers in the Gas System Planning department. Under this pressure of workload PG&E began developing new technology that would allow hundreds to thousands of pipeline simulation runs to be performed automatically as a group or "Batch".[1] PG&E now uses the Batch Analysis Tool or BAT extensively to meet workload demands, provide more thorough and accurate analysis, and provide critical system intelligence about a gas system operated frequently operating with reduced capacity due to pipeline outages to accommodate safety work. Overview of PG&E Gas System PG&E owns and operates a pipeline network that extends from both the northern California/Oregon and southern California/Arizona borders. This network of 6,750 miles of transmission and 32,000 miles of distribution pipe serves 4.4 million end use gas customers, covering two thirds of California. Pressures range from over 2000 psig on transmission to 7 inches of water column on low pressure distribution systems. Annual sendout is approximately 1 Tcf. Figure 1 provides an overview of the PG&E gas transmission system.

Abstract As described in PSIG paper 1613, PG&E's Batch Analysis Tool (BAT) allows a planning engineer to run hundreds to thousands of simulations automatically and sequentially rather than manually, one run at a time. This functionality greatly increases efficiency and allows for thorough analyses of all potential scenarios. This paper will present an actual use case study where BAT was utilized to perform over 6,500 simulation runs to gain system intelligence, optimize operations and gain customer reliability information for a multitude of operating scenarios for 256 combinations of system operating restrictions. As the location of the San Bruno pipeline incident, PG&E's Peninsula Local Transmission (LT) System has a high level of safety focus and activity.[1] In-line inspection (ILI), strength tests, pipeline replacements and automatic and remote control valving projects result in system outages throughout the year. In addition, ILI's and strength tests can result in pressure reductions, and/or pipeline sections being isolated, which can significantly reduce capacity and the ability to meet customer demand. Customer service reliability is of particular interest for pressure reductions or operating restrictions that may occur during the high demand winter heating season. This paper will present the use of the BAT in an actual study performed on the Peninsula LT system. Five major sections of the gas system faced potential operating restrictions resulting in 256 operating restriction scenarios. Each operating restriction scenario was analyzed at various temperatures, and at various operating configurations (pressure set points, cross-tie operations, interruptible non-core customer curtailments, use of portable Compressed Natural Gas (CNG) or Liquefied Natural Gas (LNG), and modeling uncertainty). In only 59 hours the BAT technology was used to complete an extensive and complex hydraulic analysis of 6,500 runs. Performing the analyses without BAT, manually a single run at a time, would have taken weeks and only yielded a fraction of the runs and system intelligence. The expansive analysis identified the most impactful operating restrictions, the best ways to operate under the restrictions, and the time of year the system would be most at risk of not meeting demand. This paper will illustrate how the power of the technology used to rapidly perform hundreds to thousands of runs provides critical system intelligence, optimize operations to maximize capacity and improve customer service reliability and fully understand operational risks when making key decisions on proceeding with the safety work.

Abstract This paper will present actual use cases of batch analysis technology using PG&E's Batch Analysis Tool (BAT) on gas distribution systems. This paper supplements the PG&E paper Pipeline Simulation Modeling Advancements - Numerous Scenarios Analyzed Simultaneously. PG&E Distribution and Transmission Planning Engineers use the BAT for a variety of functions. This paper will examine the details of specific distribution focused use cases and consider future applications for the BAT. The use cases that will be examined include: Analysis of various distribution operations required to support a transmission outage impacting the distribution system Analysis and validation of model performance Creation of system performance curves for various operating configurations System intelligence, operational optimization, and efficiency of analyses will be demonstrated. Overview of the Distribution Batch Analysis Tool Purpose As discussed in PSIG paper 1613 Pipeline Simulation Modeling Advancements - Numerous Scenarios Analyzed Simultaneously, the volume of hydraulic simulations performed by Gas System Planning has dramatically increased due to an unprecedented increase in the amount of reliability work across PG&E's service territory. This has resulted in a substantial workload increase for Planning Engineers. In order to keep pace with this increased workload and provide complete and comprehensive hydraulic analyses, PG&E's Gas System Planning has developed a Batch Analysis Tool (BAT). This is an Excel ® based tool that allows for user defined batch simulations of hydraulic models. This tool enables Planning Engineers to set up a series of simulations to be performed in an automated, sequential order rather than manually performing each simulation one at a time. This allows Gas System Planning to devote more time to evaluating hydraulic analysis results and gaining system intelligence, rather than performing simulations manually. When conducting a hydraulic analysis, multiple input parameters influence a system's performance. For example, during a system outage analysis, the system will perform differently depending on various factors including the specified ambient temperature, the system's operating pressure, valving operations, etc. If every possible combination of these parameters were to be examined, the total number of simulations would increase geometrically. Performing an analysis of this magnitude manually would require a significant amount of time. The BAT allows the Planning Engineer to examine and vary each of these parameters, creating simulations ranging from 10 runs to thousands of runs, if desired. Performing such detailed analysis is both comprehensive and time efficient.

Abstract Pipeline simulation always brings up the question of which mesh (or "knot spacing") to use. More mesh points mean improved accuracy, but also slower simulation. In addition, if the spatial mesh isn't chosen appropriately for the time step (which may itself be changing), the model can become unstable. These issues don't really have anything to do with pipelines, and it would be nice if the pipeline modeller didn't have to think about them. This is the promise of an automatically adaptive mesh. Ideally, adaptive techniques will add mesh points when and where they're needed, get rid of them when they aren't, and maintain a guarantee that the model solution is accurate to some user-specified level. Adaptive meshes are widely used in fields such as reservoir simulation[1] and pipe stress simulation[2], where 2D and 3D partial differential equations (PDEs) are solved, but haven't found wide application to the 1D+time PDEs in our pipeline models. This is because 2D and 3D PDEs often end up with so many mesh points that non-adaptive techniques just aren't feasible. In pipelines, however, if we're willing to wait a little while for the solution, we can always just throw more knots at the problem. The big question when developing an adaptive technique is how to tell where to add the mesh points. The goal is to keep the truncation error - the error introduced by the limited mesh - under specified limits. The straightforward approach to adding mesh is to compare model results between a solution with the current mesh and a denser mesh; where results differ, this suggests more mesh is needed. This involves running the model twice, once with a denser mesh than is necessary, so if a heuristic could instead be developed to estimate the needed mesh that could provide better performance. This paper will examine this two-mesh approach, comparing its performance with a fixed constant-density mesh. There are several options available for the exact algorithm used, which will be investigated. There will also be discussion of heuristics for mesh determination and the advantages and problems the author has encountered in some years of adaptive mesh use. Introduction Pipeline models solve some set of conservation equations for mass, momentum, and/or energy. These are partial differential equations and are generally complicated and nonlinear, and so must be solved numerically. In the pipeline simulation industry, the most popular approach for doing this is the method of finite differences. That means either the user or the simulation program must pick a mesh of points in space and time to use for the solution. If the mesh is chosen poorly, the user risks model instability - meaning, a solution that contains obviously nonphysical oscillations and is not usable; or, a simulator that runs very slowly; or, just a solution that looks plausible, but isn't very accurate. (This error introduced into the model solution due to limitations of the mesh, either in space or time, is called the truncation error.) Most of the commercial simulators now automate the selection of time mesh spacing (the " timestep") but it is generally still left up to the user to choose the spatial mesh. A major component of learning to use a pipeline simulation tool is learning what kind of mesh spacing to use in what circumstances.