Multiproduct Pipeline Research Articles

DeepPipe: A Multi-Stage Knowledge-Enhanced Physics-Informed Neural Network for Hydraulic Transient Simulation of Multi-Product Pipeline

In the chemical pipelining industry, owing to the high-pressure transportation process, an accurate hydraulic transient simulation tool plays a central role in preventing the slack line flow and overpressure from causing pipeline operation treacherous. Nevertheless, the current model-driven method often faces challenges in balancing computational efficiency with accuracy, and the existing data-driven models struggle to produce explainable results from the physics perspectives since insufficient theoretical principles are incorporated into the model training. Additionally, the existing physics-informed learning architecture fails to achieve a gradient-balanced training, resulting from the significant magnitude difference in outputs and multiple loss terms. Consequently, a Multi-Stage Knowledge-Enhanced Physics-Informed Neural Network (MS-KE-PINN) is proposed for the hydraulic transient simulation of multi-product pipelines. To enforce the neural network producing simulation results with high consistency to physical laws, the governing equations, boundary, and initial condition are incorporated into the training process for an efficient mesh-free simulation. Then, considering that the significant magnitude difference between outputs can easily lead to deficient performance in the gradient descent, the magnitude conversion on the outputs and the equivalent conversion of the governing equations are implemented to enhance the training effect of the neural network. Subsequently, to tackle the imbalanced gradient of multiple loss terms with fixed weights, a multi-stage hierarchical training strategy is designed to improve the approximation capacity of the neural network. Numerical simulation cases demonstrate a better approximation function of the proposed model than the state-of-art models, while the mean absolute percentage errors yielded by MS-KE-PINN are reduced by 77.4 %, 88.7 %, and 87.8 % in three simulation operation conditions for pressure prediction. Furthermore, experimental investigations from a real-world multi-product pipeline suggest that the proposed model can still draw accurate simulation results even under complex and dynamic hydraulic transient scenarios in practice, with root mean squared errors reduced by 94.8 % and 80 % than that of the physics-informed neural network. To this end, the proposed model can conduct accurate and effective hydraulic transient analysis, thus ensuring the safe operation of the pipeline.

Pipe sharing: A capacity allocation method for incorporating economy and fairness in the multiproduct pipeline

To guarantee fair access to pipeline facilities, pipe companies around the world have promoted market-based reform measures that are subject to governmental regulation. The pipe company should allocate transport capacity to shippers based on their nominations and then make schedules that align with the operational process. The reasonable mechanism is crucial for allocation results, affecting the carrier's revenue and shippers' fairness. Aiming at this issue, this paper puts forward an allocation mechanism of combining priority rules with historical data based on summarizing existing mechanisms. A multi-objective optimization model for capacity allocation that simultaneously considers economic and fairness metrics is developed. Finally, the efficient frontiers describing the economy-fairness trade-off are generated through using ε-constraint method, which can be selected by the pipe company based on its preference. Using data from a real pipeline as a case study, the solutions under different mechanisms are quantitatively evaluated. The results show that different mechanisms are applicable differently. Compared with other existing mechanisms, the proposed mechanism brings the pipe company the maximum cost of 1665.98 × 104CNY. On the basis of realizing the fair allocation, the proposed method ensures the pipe company to obtain the maximum transport cost to a certain extent as well.

Efficient Pump Scheduling for Large‐Scale Multiproduct Pipelines Using Deep Reinforcement Learning

ABSTRACTOptimizing pump scheduling in multiproduct pipelines can significantly reduce energy consumption and carbon emissions. For pump scheduling in multiproduct pipelines, to describe hydraulic losses more accurately, the model needs to adopt shorter discrete time intervals, which will lead to longer decision‐making time. It combined with the large solution space of large‐scale pipelines, will lead to low solution efficiency with dynamic programming methods and poor solution quality with heuristic optimization algorithms. Given that, this article develops a multiproduct refined oil transmission simulation system and employs the enhanced Proximal Policy Optimization (PPO) algorithm with action space shaping trick to optimize pump scheduling for large‐scale multiproduct pipelines. The method of converting discrete action space to multi‐discrete action space through action shaping can address PPO's low convergence efficiency issue resulting from the large discrete action space challenge common in large‐scale multiproduct pipelines. The experimental results indicate that the proposed method, that is, PPO algorithm with multidiscrete action space exhibits significant advantages in terms of efficiency and robustness in large‐scale pipelines compared to mainstream methods for pump scheduling such as dynamic programming (DP), genetic algorithms (GA), and ant colony optimization (ACO). Furthermore, we demonstrate the effectiveness of action space shaping in large‐scale pipelines from the perspectives of exploration and exploitation.

Online detection of mixed oil interface in multi-product pipeline using near-infrared spectroscopy

To ensure the quality of refined oil products of sequential transportation, it is necessary to monitor and judge the interfaces between pure oils and mixed oil for guiding accurate mixed oil cutting. In this study, near-infrared spectroscopy (NIRS) was used to investigate the change trend of oil concentration and thus judge the interfaces. The methods of standard deviation of continuous spectra, principal component analysis, difference spectrum, and vector angle θ were used to analyze the online oil NIR spectra to detect the interfaces, respectively. The results show that θ of the spectra pretreated by the combination of airPLS and 1st derivative is the best for reflecting the concentration change trend of mixed oil and judging the interfaces. The interface between the forward pure oil and mixed oil can be judged by a fixed θ threshold, and the interface between the rear pure oil and mixed oil can be judged by a fixed first-order derivative value of the five-parameter logistic fitted curve of θ. This method is not affected by the interference and oil composition change of different batches and does not require oil properties and pipeline parameters, thus it has a good prospect in the oil pipeline transportation field.

A 1+ Mechanism Model for Predicting the Mixed-Oil Concentration in Multiproduct Pipelines

Summary Petroleum products are frequently transported successively through the same multiproduct pipeline. Due to turbulent and convective diffusion mass transfer, two adjacent oils will mix with each other, forming a mixed-oil segment. Accurate and rapid prediction of mixed-oil concentration is crucial for the precise management of mixed-oil segments. Conventional 1D modeling methods exhibit shortcomings in accurately representing the asymmetric distribution characteristics of mixed-oil concentration curves, and high-dimensional models are not practically applied due to their prohibitive computational time costs. Building on the 1D model framework, this paper proposes a “1+” mechanism model by considering the convective mass transfer behavior between the turbulent core region and the laminar boundary layer, and new governing equations and corresponding numerical solution methods are also introduced. Simulation experiments affirm the ability of the new model to characterize the asymmetric distribution features of mixed-oil concentration curves, along with its high computational efficiency in engineering applications. This is demonstrated by the computational time of approximately 30 seconds for simulating a pipeline of 300 km in length (Δx = 10 m, Δt = 1 second, CPU: i5-12500H, RAM: 16 GB). When applied to pipelines in industrial scenarios, the new model is shown to accurately predict the distribution of mixed-oil concentration curves. The research findings are significantly beneficial in assisting field personnel to gain advanced insights into the mixed-oil concentration distribution at the station, enabling timely and well-informed strategies for handling mixed-oil segment, thereby enhancing the operational efficiency of multiproduct pipelines.

Two-stage optimization model for scheduling multiproduct pipeline network with multi-source and multi-terminal

The multiproduct pipeline network with multi-source and multi-terminal encompasses multi-point injection and multi-point distribution. Upon reorganization of batches at transfer depots, the intricacies of precise batch tracking become increasingly complex, thereby rendering the optimization of multiproduct pipeline network scheduling particularly arduous. In this paper, we introduce a dynamic index batch numbering method. By delineating the boundaries of intermediate injection points, the issue is bifurcated into two sub-problems: the optimization of batch planning for multiproduct pipeline with single-source and multi-terminal and the optimization of resource allocation at intermediate oil depots. The proposed algorithm is validated on two genuine multiproduct pipeline networks in China, demonstrating its capability to generate feasible solutions within a computational constraint of 18,000 CPU seconds. The findings reveal that this methodology holistically considers the entire pipeline network system, encompassing upstream oil intake, downstream oil supply, and the tank capacities of transfer depots, thereby establishing a coherent link between pipeline scheduling and depot planning. This paper may serve as a reference for the automated formulation of multiproduct pipeline network scheduling strategies.

Soft sensor development for mixed oil interface tracking in multi-product pipelines based on knowledge-informed semi-supervised Variational Bayesian Gaussian mixture regression

Sequential transportation of petroleum products in multi-product pipelines often lead to occurrence of mixed oil. The prediction of the arrival time of the mixed oil interface constitutes crucial data for the scheduling of treatment actions. However, existing soft sensors, like Gaussian mixture regression (GMR), may face challenges due to limited size of labeled data and numerical issues, leading to performance degradation or even training failure. To tackle these issues, we propose a Knowledge-informed Semi-supervised Variational Bayesian Gaussian mixture model (KI-SSVBGMR). It employs a semi-supervised fully Bayesian structure designed to address the constraints arising from the potential matrix singularity issues and shortage of labeled samples. Subsequently, we determine the crucial regression variable and establish its prior distribution based on industrial knowledge to improve model generalization. Finally, a learning procedure grounded in the Variational Inference algorithm is developed to train the KI-SSVBGMR. Through case studies, including numerical examples and real industrial datasets, our method demonstrates the effectiveness and reliability of the proposed soft sensor development method. This research can aid operators in improving mixed oil section management and provides valuable insights for integrating machine learning with industrial knowledge.

Pipeline sharing: Optimal design of downstream green ammonia supply systems integrating with multi-product pipelines

Green ammonia is a promising energy storage form with extremely high efficiency of hydrogen storage. Although the construction of green ammonia production facilities is rapidly progressing, there is still a lack of comprehensive consideration of the optimal design for supply systems. This paper prospectively explores a framework for the ammonia supply system under four scenarios considering multimodal transportation. Different mathematical models are developed, which incorporate special constraints such as pipe diameter, broken loops, etc., aiming at minimizing costs of transportation, construction and carbon treatment. In this framework, the economic, energy and environmental performance is evaluated to optimize the design of the supply system. The case study in China shows that different transportation modes can reduce economic costs by 60.93∼77.45 %, energy consumption by 18.4 ∼67.7 ktce, and carbon emissions by 23.3∼60.7 kt per year. A further novel finding is that the simultaneous usage of new liquid ammonia pipelines and multi-product pipelines for supply systems proves to be the most effective transportation mode. Sensitivity analyses are also conducted on regional production capacity expansion, liquid ammonia pipeline planning, and multi-product pipeline utilization. The results confirm that the scheme is stable under certain conditions. Finally, some recommendations are provided for the development of green ammonia industry.

Pipeline sharing: Boosting multi-product pipeline transport biofuels in the shift to low-carbon energy

Under the dual pressures of oil depletion and low carbon transition, biofuels are considered effective alternatives for fuel oil due to their renewable nature and environmental friendliness. However, oil demand remains high, which indicates that biofuels and fuel oil will coexist for a long time to come. While pipelines offer an efficient means of transporting liquid fuels, the initial investment required is substantial. Therefore, this paper proposes a comprehensive framework that focuses on sharing already existing multi-product pipelines with biofuels, thus evaluating the pipeline remaining potential to transport biofuels. The core component of this framework is the evaluation method for determining the multi-product pipeline remaining capacity. We coupled fuel oil consignment contracts into the remaining capacity evaluation method and developed a model for evaluating the multi-product pipeline transport biofuels potential, aiming to maximize biofuels pipeline transport volume. The objective of this model is to optimize the transportation volume of biofuels through pipelines. Various constraints are taken into account, such as product batch movement, prohibited sequences, operational constraints, delivery time frames, and inventory levels. Moreover, this study evaluates and discusses the benefits of biofuels transportation capacity using eight real-world multi-product pipeline systems without impacting fuel oil consignment contracts. The findings indicate that pipelines are capable of transporting 49,849 m3 of biofuels per week, increasing revenue by 1.78 million CNY. Simultaneously, the low energy consumption for pipeline transportation can lead to a weekly reduction of 34.6 tons in carbon emissions. The evaluation model anticipatively provides the ability to respond to future market developments.

The Component Displacement Process of Two Miscible but Dissimilar Fluids Transported Sequentially in a Multiproduct Pipeline

The Component Displacement Process of Two Miscible but Dissimilar Fluids Transported Sequentially in a Multiproduct Pipeline

Scheduling of Straight Multiproduct Pipelines Considering the Contamination Control

Scheduling of Straight Multiproduct Pipelines Considering the Contamination Control

Knowledge-informed Variational Bayesian Gaussian mixture regression model for predicting mixed oil length

Precise prediction of mixed oil length is a critical task in the multi-product pipeline. Current predictive models may not achieve desirable performance as they neglected the multimode characteristics of the data, which has been alleviated by using the finite mixture model (FMM) to identify the different modes and develop localized predictive models while it faces the challenge of assigning a suitable number of components. Variational Bayesian Gaussian mixture regression model (VBGMR) may remedy these deficiencies but the linear assumption between variables can result in unsatisfactory performance, and it may also not handle the model selection well since all input variables are forced to participate in the mode identification process, even if some are not important. To address such issues, this article proposes a new modeling approach named as the Knowledge-informed Variational Bayesian Gaussian mixture regression model (KIVBGMR) by incorporating the conventional VBGMR with the guidance of the domain knowledge. Subsequently, the learning procedure for the KIVBGMR based on the Variational Inference method is developed. The performance of the KIVBGMR modeling method is evaluated based on two case studies including a numerical example and a real-life industrial dataset, which demonstrates the performance superiority of the proposed method.

Pipeline sharing: A supplementary strategy to enhance 3E performance of multi-product pipelines in China

Single-rate pipeline prices have resulted in low competitive flexibility and have led to high costs and consumption in China's refined oil logistics system. This study aims to propose a framework for multi-product pipelines supplementary pricing strategy that creates a win-win situation for both shippers and pipeline carriers while quantifying economic, energy and environmental performance. Based on the objective of cost minimization, a lower limit constraint on pipeline transportation quantity is added, and a discount coefficient is introduced for pipeline prices, thus improving the transportation optimization model. Moreover, this paper develops an implementation approach for a supplementary pricing strategy that can ensure that the discounted price is not lower than the sum of cost and minimum profit. This framework usability is confirmed using a Chinese oil company as an example. The results show that the optimal scheme can increase pipeline revenue by ¥79.8 million while reduce shipper’s logistics costs by ¥1.68 million during the practice cycle. Meanwhile, this work reduces energy consumption by 1.37 tce, CO2 emission by 444 tons, SO2 emission by 5.65 tons, and NOX emission by 7.07 million tons for the logistics system. Compared with past studies, this study is more economical or more in line with market-oriented behavior.

Deeppipe: An intelligent framework for predicting mixed oil concentration in multi-product pipeline

Accurately predicting mixed oil concentration distribution exerts a core effect on the optimization of pipelines and the quality of oil. Due to the neglect of mechanism features, high-dimensional complex feature correlations, and insufficient feature information on small batch data, the current methods cannot predict mixed oil concentration accurately. This work proposes a hybrid intelligent framework to provide an accurate and effective monitoring tool for mixed oil concentration of multi-product pipelines. In the proposed framework, the development mechanism of mixed oil is analyzed to select and reconstruct holistic features to explore the influencing mechanism of mixed oil concentration. Then, a parameterization and nonlinear transformation module is designed to acquire the accurate and concise representation of mixed oil concentration, thus decreasing the complexity of feature space and promoting the approximating ability of the prediction model. Eventually, a novel virtual samples generation module is established to obtain high-quality samples of mixed oil concentration, aiming to extract more comprehensive correlations of feature variables and improve the prediction performance. Cases from real-world multi-product pipelines suggest more accurate prediction results of mixed oil concentration compared to other advanced methods, with RMSE and R2 being 0.0500 and 0.9688. Furthermore, it is also proved that acquiring more holistic and accurate feature variables of mixed oil development and fully exploring comprehensive correlations between feature variables are crucial for the performance enhancement of the prediction model.

A multiproduct gasoline supply chain with product standardization and postponement strategy

• The gasoline supply chain proposed in the 2013 Mexican energy reform is inviable. • A supply chain based on gasoline standardization, collaboration and postponement. • A multi-product pipeline inventory-transport problem. • A Mixed-Integer Nonlinear Programming Problem global solution methodology. • A real-life study case.

Energy saving and consumption reduction in the transportation of petroleum products: A pipeline pricing optimization perspective

Pipeline transportation is a low-energy and economical mode of transporting petroleum products in the downstream supply chain, however, there is almost no theoretical research on its pricing strategies. The unreasonable pricing strategy has resulted in low utilization of multi-product pipeline capacity as well as high energy consumption of petroleum products transportation. Therefore, this paper aims to improve pipeline turnover and promote the low-carbon transportation market from the perspective of pipeline pricing optimization. We propose an integrated framework for multi-product pipelines that couples pricing strategy and logistics optimization model. This framework simulates the pricing behavior of the pipeline carrier and the corresponding logistics planning behavior of the oil shipper. We apply the framwork to 10 pipeline pricing schemes for two regions in China with different logistics structures, and analyze the economic and environmental benefits of the new strategy. The results show that the well-performing scheme can increase pipeline carriers' revenue by 11.41 million CNY per month, significantly improve the competitive advantage of long-distance pipelines, and reduce energy consumption by 272 tce. Based on these findings, we provide recommendations for policymakers at four levels. In conclusion, the new pricing strategy will help reverse the disadvantageous situation of the pipeline in the competitive market and promote energy conservation in the petroleum products logistics industry.

Liquid pipeline network modeling with performance parameters sensitivity analysis due to its techno-commercial variables

After analyzing existing pipeline-network design solutions in literature, the authors have developed a novel framework model that considers important technical, financial, and environmental factors such as terrain profile, thermal impact, dynamic pump efficiency, cost escalation, tariffs, project phasing, depreciation, and carbon emissions. This comprehensive model not only prioritizes cost optimization for performance measurement but also takes into account tariffs and carbon emissions. To validate the model, the authors conduct a case study on a China multiproduct pipeline and perform sensitivity analysis. The technical model is accurate to 1.1%, except for upstream pressure at the end receiving station. Sensitivity analysis reveals that incorrect judgment of elevation profile, volume escalation, and product temperature variables can render the outcome of pipeline design infeasible. Moreover, neglecting financial components, such as lines fill cost and tankage cost, can cause a variation of 23% in the performance parameters, leading to erroneous decision-making in Pipeline Network Configuration (PNC) design and operations. Lastly, the authors discuss the significance of considering tariff and carbon-emissions performance parameters during design optimization.

Pipeline sharing: Potential capacity analysis of biofuel transportation through existing pipelines

In response to resource depletion and tightening carbon emissions policies, the product oil is gradually being substituted by biofuels and faced with falling demand. A large number of studies have been done on the optimization of the biofuel supply chain, but few of them have considered the existing petroleum pipelines for policy reasons. Market-based reforms in the oil and gas industry have provided the possibility of pipeline sharing, whereby pipeline owners can use the remaining delivery capacity to transport other liquid fuels that are allowed into the pipeline, such as biofuels, thereby increasing pipeline benefits and reducing biofuel logistics costs. This paper develops a logistics optimization model considering multi-product pipeline scheduling from the perspective of biofuel suppliers and proposes a pipe-rail transportation mode to assess the optimization potential of logistics costs. Finally, this paper verifies the feasibility of the proposed approach using a logistics system in a region of China as a practical case study and conducts a sensitivity analysis on whether changes in the demand for product oil and biofuels affect the choice of pipeline opening location, concluding that there is no effect within a 4-fold increase in biofuel production and a 0.78-fold decrease in product oil demand.

Development of machine learning model for determination of contamination length in a multi-product pipeline

Batch transfer results in contamination over the length of travel of the fluids in product pipelines. Mathematical models have been in use. Machine learning with Python, due to higher efficiency was applied to determine contamination length. Six models were developed and the best with an accuracy of 97.4% and RMSE score of 262.5 was developed. It predicts with higher precision and also accurately ranks the input variables in order of their influence on transmix length. The distance of travel had the highest influence on the amount of contamination in a pipeline, followed by Reynolds number and pipe diameter.

Pipeline sharing: Remaining capacity estimation of multiproduct pipelines

Accurate estimation of remaining capacity is crucial for pipeline companies to improve their service quality and economic benefits. Considering the delivery of multiple batches in long-distance pipelines, the remaining capacity is affected by technical capacity, delivery time, booked capacity, injection location, transmission and so on. Taking the maximum revenue as the objective, this paper develops a mathematical model for the multiproduct pipeline containing dual-purpose nodes to obtain its optimal remaining capacity at different injection nodes during different periods. The proposed model is validated by a real-world pipeline in China, and the sensitivity analysis is also conducted to identify the driving factors of the optimal remaining capacity. The comparison reveals that the tightness of delivery time has a major impact on the optimal remaining capacity, especially at downstream nodes of the pipeline. Finally, relevant suggestions are put forward based on the results to help pipeline operators make decisions.