Decision Models for Bulk Energy Transportation Networks

我们推测，如果向行业参与者提供一个基于经验的综合能源系统模型，允许对生产、储存、运输、转换和交付替代方案进行全面评估，那么美国现有的能源系统可以显著提高效率、可靠性和环境安全。在美国，超过76%的电能由煤、天然气或水提供。这三种原始能源形式，连同电力，都有一个共同的特点，那就是它们可以通过运输系统从它们的生产地大量运输到它们的使用地。煤炭主要通过火车和驳船运输；天然气通过管道运输；水通过河流和水库系统运输；电力通过输电线路运输。美国许多不同负荷中心从原始能源到发电厂再到配电所的移动模式是由许多不同组织组成的复杂决策过程决定的。因此，我们的研究目标是开发决策模型来解决三个相关领域的问题：(1)什么样的能量流动模式将显著改善能源系统的性能？为了实现这些改进，需要进行哪些生产和/或运输方面的操作改变？(2)基础设施存在哪些薄弱环节？哪些基础设施改进将实现最大的性能收益？(3)我们可以多好地预测市场设计变化对能源系统性能的影响？我们将开发两种类型的模型来回答这些问题。结构优化模型将实现对能源系统、其物理基础设施及其相关市场流动的中期(1个月至2年)模拟。行为模型将重点放在覆盖结构模型的市场网络上。市场参与者被建模为战略能源交易员，他们有能力自主行动，并从之前决策的影响中学习。这两个模型通过交易商-代理人产生的能源价格/数量/地点/时间出价和要约相互作用。我们将使用参与式模型开发过程，在此过程中，我们反复循环进行实地工作、模型设计和开发以及计算实验。尤其令人感兴趣的是制约和塑造原燃料生产和运输公司以及从事发电、输电和配电的公司的决策过程的结构性和社会因素。这些知识将用于指导结构和市场模型的发展，从而提高这些模型提供准确系统性能评估的能力。因此，我们将更好地了解影响系统参与者决策过程的社会和结构性因素，以及系统参与者采用新的结构和程序来提高能源传输系统效率的意愿。本研究建立的模型和进行的分析将使系统地检查全国大宗能源生产和运输决策，从而揭示具有非常大的国民经济影响的运营和设施投资效率。此外，它将暴露能源系统的脆弱性，导致能源系统设计和分析方法，该方法考虑到不确定性和中断，以指导可靠性增强和市场设计。这些努力将导致在组织和认知层面确定能源系统人类决策过程的特征，在与影响国家关键基础设施的高压力情况有关的决策中扩展人类动力学知识，同时加深对高级网络优化和软件代理之间的实际和概念关系的理解。这项工作的最终结果将包括一套健全而稳健的模型和伴随的建模方法，用于研究、分析和设计综合能源系统及其市场。这种公开领域、开放源码的工具和方法将对刺激和鼓励大学在这一领域进一步开展研究和教育产生重大影响，以新的方式将不同的领域(电力系统、社会学、决策学和经济学)联系起来，有助于丰富每个领域的教育和研究。

We conjecture that the existing U.S. energy system could be operated with significantly increased efficiency, reliability, and environmental safety if industry participants were provided with an empirically based model of an integrated energy system permitting comprehensive assessments of production, storage, transportation, conversion, and delivery alternatives. Over 76% of electric energy in the U.S. is supplied by coal, gas, or water. These three raw energy forms, together with electricity, have the common characteristic that they can be moved in bulk quantities via a transportation system from their source of production to where they are used. Coal is mainly moved by train and barge; gas by pipelines; water by rivers and reservoir systems; and electricity by transmission lines. The pattern of movement from raw energy source to power plants to electric distribution substations of the many different U.S. load centers is determined by a complicated decision-making process comprised of many different organizations. Our research objective, therefore, is to develop decision models to address three related areas of questions: (1) What energy flow patterns would yield significantly improved energy system performance? What operational production and/or transportation changes need to be made to realize these improvements? (2) What infrastructure weaknesses exist? What infrastructure enhancements would realize the most performance benefit? (3) How well can we predict the influence of market design changes on energy system performance? We will develop two types of models to answer these questions. A structural-optimization model will enable mid-term (1 month to 2 years) simulation of the energy systems, their physical infrastructure, and their associated market flows. A behavioral model will focus on the market network overlaying the structural model. Market participants are modeled as strategic energy traders having the ability to act autonomously and learn from the influence of their previous decisions. These two models interact through energy price/quantity/location/time bids and offers generated by the trader-agents. We will use a participatory model development process in which we repeatedly loop through fieldwork, model design and development, and computational experiments. Of particular interest are structural and social factors constraining and shaping the decision-making processes of raw-fuel production and transportation firms as well as firms engaged in electricity generation, transmission, and distribution. This knowledge will be used to guide the development of the structural and market models and hence to improve the ability of these models to provide accurate system performance assessments. We will thereby gain better understanding of the social and structural factors affecting the decision-making processes of system participants and the willingness of system participants to adopt new structures and procedures to improve the efficiency of the energy transmission system.The modeling developed and analysis performed in this research will enable systematic examination of nationwide bulk energy production and transportation decisions and consequently reveal efficiencies in operational and facility investment having very large national economic impact. In addition, it will expose energy system vulnerabilities, leading to an energy system design and analysis approach that accounts for uncertainties and disruptions to steer reliability enhancement and market design. These efforts will result in characterization of energy system human decision processes at organizational and cognitive levels, extending knowledge of human dynamics in decision-making related to high-stress situations affecting a national critical infrastructure, while deepening the understanding of practical and conceptual relations between advanced network optimization and software agents. The culmination of this work will include a sound and robust suite of models and accompanying modeling methods for study, analysis, and design of the integrated energy system, together with its markets. Such public domain, open-source availability of tools and methods will have significant effect on stimulating and encouraging further university-based research and education in this area, bridging separate fields (power systems, sociology, decision-science, and economics) in new ways that will contribute to enriching education and research in each of the areas.