CDI-Type II: Energy Policy, Investment, and Pricing Analysis Driven by Computational Steering

[1029337]对美国的担忧美国对化石燃料的依赖增加了减少这种依赖的动力。增加可再生能源、能源存储、插电式混合动力汽车（phev）和电动汽车（ev）通常被认为是解决方案的一部分。然而，这些技术的广泛应用将对美国电网产生深远的影响。因此，了解插电式混合动力车、电动汽车、可再生能源、储能、分布式发电和电网之间相互作用的经济和政策问题是很重要的。允许在不同假设集下模拟策略的集成模型对于理解此类系统的复杂性和紧急行为至关重要。能源、环境和国家安全问题的交叉涉及一系列复杂的权衡，包括初始成本、当地空气质量、气候变化、技术和能源独立。如果不了解不同规模和不同能源生产部门之间的相互作用，就只能分析部分和过于简单的解决方案。该项目将创建一个下一代电力系统的综合模型，将个人和电网面临的经济决策整合在一起。称为能源定价和政策综合计算系统（ICS-EPP）。主要项目成果将是ICS-EPP模型，该模型包括一个典型的电网（如美国的电网）和个体的自主子模型，以有效地模拟在不同经济政策和技术选择下这种系统演变的不同情景。ICS-EPP的设计旨在解决一些重要问题，例如将插电式混合动力车和电动汽车加入电网的成本，车辆对电网服务的价值，或不同定价结构对能源使用和投资的影响。ICS-EPP的全面性使其能够检查各种行动的副作用：例如，排放量的变化和对发电能力的需求。更重要的是，随着决策者就减少温室气体排放和减少对化石燃料依赖的各种机制进行辩论，他们将对各种政策选择的成本和效果有更准确的认识。州和地方政府可以研究激励不同技术的价值，如插电式混合动力充电站、家用太阳能电池板和/或集成热电机组。该模型将允许政策制定者和分析人员在一个新的细节水平上对家庭、能源技术和电力系统的相互作用进行虚拟实验。该项目利用了广泛的合作，包括工程师，经济学家，计算科学家，以及汽车和电力行业的领导者。智力优势：该项目依靠计算导向将家庭模型和电网模型结合并求解成一个大规模的基于智能体的模型。由于适应市场的微妙之处和解决这样一个模型所需的规模和计算能力，这种建模还没有尝试过。这项工作的前提是，计算建模和模拟可以作为评估不同经济政策潜在影响的有效手段。这将需要在若干领域取得进展：(a)将基于不同模拟方法的不同系统的模型整合为一个综合整体的能力；(b)从可与定性目标相匹配的大型模拟结果中获取信息的技术；(c)在当前流行的高性能计算系统上在多个时间尺度上运行大型异构模型。更广泛的影响：该项目的成功将推动能源政策分析的范式转变，因为决策者可以从多个维度（价格；温室气体、硫氧化物和氮氧化物排放；石油消费；家庭支出等）检查针对或影响能源的政策的结果，例如购买插电式混合动力汽车或充电基础设施的税收抵免，温室气体排放税或上限，或将插电式混合动力汽车与太阳能和风能发电相结合的协同效应。此外，还将创建一个可以应用于其他领域的低级计算指导工具包。研究也将以跨学科课程的形式整合到经济学、工程学和计算机科学课程中。项目官员：Bruce HamiltonCBET部门工程理事会代表CDI工作组日期：2010年8月9日

1029337SioshansiAbstractConcerns about the U.S.'s reliance on fossil fuels have increased the incentive to reduce this dependence. Increased renewable energy, energy storage, and plug-in hybrid vehicles (PHEVs) and electric vehicles (EVs) are often suggested as parts of the solution. However, widespread use of these technologies will have a profound effect on the U.S. power grid. It is therefore important to understand economic and policy questions regarding the interplay of PHEVs, EVs, renewable energy, energy storage, distributed generation, and the grid. Integrated models that allow the simulation of policies under different sets of assumptions are crucial to understanding the complexity and emergent behaviors of such systems. The intersection of energy, environmental, and national security issues involve a complex set of trade-offs among initial cost, local air quality, climate change, technology, and energy independence. Without understanding the interactions of the different scales and different sectors of energy production, only partial and overly simplistic solutions can be analyzed. This project will create a comprehensive model of next-generation power systems that integrates the economic decisions faced by both individuals and the power grid?called the Integrated Computational System for Energy Pricing and Policy (ICS-EPP). The main project outcome will be the ICS-EPP model that includes a typical power grid (such as that of the U.S.) and autonomous sub-models of individuals to efficiently simulate different scenarios of the evolution of such a system under different economic policies and technology options. The ICS-EPP will be designed so that important questions can be addressed, such as the cost of adding PHEVs and EVs to the grid, the value of vehicle to grid services, or the impacts of different pricing structures on energy use and investment. The comprehensive nature of ICS-EPP permits examination of the side effects of various actions: changes in emissions and demand for generating capacity, for example. More important, as policy-makers debate various mechanisms to reduce GHG emissions and reduce dependence on fossil fuels, they will have available a more accurate view of the cost and efficacy of various policy options. State and local governments can examine the value of incentivizing different technologies such as PHEV charging stations, household solar panels, and/or integrated heat and power units. This model will permit policy makers and analysts to conduct virtual experiments of the interaction of households, energy technologies, and the power system at a new level of detail. The project leverages a broad collaboration that includes engineers, economists, computational scientists, as well as automotive and electric power industry leaders.Intellectual Merit:This project relies on computational steering to knit together and solve models of households and the power grid into a large-scale agent-based model. Such modeling has not been attempted, due to the size and computational power needed to accommodate the subtleties of the markets and solve such a model. The work is based on the premise that computational modeling and simulation can serve as an effective means of evaluating the potential effects of different economic policies. This will require advances in a number of areas: (a) the ability to integrate models of diverse systems based on different simulation methodologies into an integrated whole; (b) techniques for information from large simulation results that can be matched against qualitative objectives; and (c) running large, heterogeneous models at multiple time scales on currently prevalent High Performance Computing systems.Broader Impact:The success of this project will enable paradigm-shifting advances in energy policy analysis, as decisionmakers can examine the results, in multiple dimensions (price; GHG, SOx, and NOx emissions; petroleum consumption; household expenditures; etc.), of policies that target or affect energy, such as tax credits for PHEV purchase or charging infrastructure, GHG emissions taxes or caps, or the synergies of integrating PHEVs with solar and wind generation. Also, a low-level computational steering toolkit that can be applied to other domain areas will be created. Research will also be integrated into economics, engineering, and computer science curricula in the form of interdisciplinary courses.Program Officer: Bruce HamiltonCBET DivisionDirectorate for EngineeringOn Behalf of CDI Working GroupDate: August 9, 2010