Energy Foundation Wind Integration Meeting U.S. Wind Integration Studies and Doing a Wind Integration Study 美国风电并网研究 及 研究方法 Kevin Porter 凯文 波特 Exeter Associates 公司 Energy Foundation Wind Integration Meeting 并网经验介绍会 能源基金会 July 2, 2010

Purpose of Presentation 介绍会的目的 Determine whether there is interest in learning more about how to do a wind integration study. 了解大家是否有兴趣进一步学习开展风电并网问题研究的研究方法。 Determine whether there is interest in holding a workshop on wind integration study methods in Beijing in December. 了解是否需要于今年12月在北京再组织一个风电并网研究方法介绍会 Consider interest in conducting a wind integration study in China. 考虑一下在中国开展风电并网研究的意义 Exeter Associates, Inc.

Overview of U.S. Wind Integration Studies 美国风电并网研究概况 Exeter Associates, Inc.

Overview of U.S. Wind Integration Studies 美国风电并网研究概况 Several utilities, states and regional transmission organization have conducted wind integration studies, generally before large-scale wind development has occurred Over a dozen such wind integration studies since the late 1990s, and more are planned Actual operating experience has been limited, therefore studies have been prospective – modeling a potential future power system with certain amounts of wind generation Studies focused on reliability, operational and economic impacts of higher levels of wind penetration Every U.S. wind integration study is different, but some common themes are emerging Studies have evolved from whether it is possible to incorporate wind to how and at what estimated cost As experience is gained, studies are starting to focus on potential solutions to integrating higher levels of wind generation 总的来说，在大规模并网实际发生之前，美国的一些电力公司、州及区域输电组织已经开展了风电并网问题的研究。 上世纪90年代至今这样的研究做了有十几个，目前计划开展更多此类研究。 由于实际运行经验的局限，前瞻性研究变得更为重要---模拟出接入一定规模风电情景下的未来的电力系统 研究基本聚焦在大规模接入风电后系统的可靠性问题、运行影响和经济影响。 每个研究都有所区别，但更多的共同关注问题已经开始显现。 研究的重心已经从原来研究是否可能接入这么多风电，转为研究如何以及在多大经济代价下接入这些风电。 由于已经有了一些经验基础，研究重点已经开始集中在去寻找接入更高比例风电的潜在解决办法。 Exeter Associates, Inc.

Integration Cost Estimates 并网成本估算 Wind integration studies have found cost will generally be under $5.00/MWh for wind capacity penetrations up to 20%; Most of these costs are from unit commitment; Pacific Northwest wind studies have tended to show costs up to $10/MWh – No sub-hourly scheduling or markets Greater reliance on regulation versus spinning or non-spinning reserves Tends to increase costs U.S. has more flexible power system than China 研究发现，风电接入比例在20%左右或以下的情况下，并网成本一般在每千度电5美元左右（0.5美分/千瓦时，即每度电人民币3分4厘左右）。 机组组合成本占这些成本的大部分； “西北太平洋”的研究曾给出一个接近每千度电10美元左右的并网成本（1美分/千瓦时，即每度电人民币6分8厘左右）– 没有小时内调度或相应市场 更多地依赖调节而非旋转备用或非旋转备用 倾向于提高成本 美国的灵活性电源多于中国 Exeter Associates, Inc.

Wind Integration Costs Dependent on Several Factors 影响风电并网成本的几个因素 Size of the balancing area Large balancing area makes it easier to integrate wind, as there will be a deeper stock of generation to draw from Subhourly markets access the flexibility inherent in many conventional generators and reduces reliance on regulation in following wind’s variability Resource mix A resource mix with more flexibility in ramping up and down and operating at different dispatch points will make wind integration easier Depth and type of ancillary services A well-functioning and deep market for ancillary services (present in most regional transmission organizations) will make it easier and less expensive to integrate wind The geographic concentration of wind projects Greater spatial diversity of wind projects can lessen the variability in wind output and lower wind integration costs 平衡区域的大小 大平衡区接入风电更容易，因为可以利用发电容量大的优势 时间间隔小于1小时的电力市场可以利用常规电源的灵活性，减少对调节风电波动的依赖。 资源组合 资源组合具有更好的爬坡（上下）灵活性，可以根据不同的调度要求运行，从而减少风电接入的难度。 辅助服务的深度和类型 完善、繁荣的辅助服务市场(多数由区域输电组织提供) 可以减少风电接入的难度和成本 风电项目地理分布的密集度 风电项目在空间上的分散式分布可以减少风电的出力波动并降低并网成本 Exeter Associates, Inc.

Key Results from Major Wind Integration Cost Studies 主要风电并网成本研究的主要结论 Date 时间 Study 研究 Wind Capacity Penetration 风电装机容量 整体占比 Integration Cost ($/MWh) 并网成本(美元$/千度电) Regulation 调节 Load Following 负荷跟踪 . Unit Commit 机组组合 Gas Supply 天然气供应 TOTAL 总计 2003 Xcel-UWIG 3.5% 0.41 1.44 Na 1.85 We Energies 29% 1.02 0.15 1.75 2.92 2004 Xcel-MNDOC 15% 0.23 na 4.37 4.60 2005 PacifiCorp-2004 11% 1.48 3.16 4.64 2006 Calif. (multi-year)* 4% 0.45 trace Xcel-PSCo 0.20 3.32 1.45 4.97 MN-MISO* 31% 4.41 2007 Puget Sound Energy 12% 6.94 Pub. Service 0.37 2.65 1.06 4.08 Avista Utilities 30% 1.43 4.40 3.00 8.84 Idaho Power 20% 7.92 PacifiCorp-2007 18% 1.10 4.00 5.10 2008 Xcel-PSCo* 8.56 2009 Bonneville (BPA)+ 36% 0.22 1.14 5.70 PacifiCorp-2009++ 22% 11.85 Portland Gen. Elect. 27% 11.75 2010 EWITS+++ 48% 1.61 4.54 Nebraska 63% * See notes from 2008 Wind Technologies Report详见2008风电技术报告注解 + Costs in $/MWh assume 31% capacity factor. The aside from regulation and following reserves, the cost of BPA’s imbalance reserves are $4.33/MWh. ++ This integration cost assumes a $45/ton CO2 tax. With a lower CO2 tax of $/ton, the integration cost decreases to $9.96/MWh +++ Unit commitment cost is listed in EWITS as the cost of day-ahead wind forecast error, remaining integration costs are for shorter term reserves Exeter Associates, Inc.

Exeter Associates, Inc.

Solutions: Implement a Wind Forecasting System 解决办法：应用风电预测系统 In general, wind forecasting is vital once wind penetrations by energy reaches 10% Lower if balancing area is small or if resource mix is not overly flexible Although wind forecasting is not perfect, today’s state-of-the-art wind forecasts will be helpful to grid operators Potential large savings in unit commitment and fuel consumption costs Promise for future performance improvements in wind forecasting Several grid operators in the United States are implementing wind forecasting Evolution towards central wind forecasting and away from decentralized wind forecasting 一般来讲，一旦风电在电力供应中上到10%，预测就变得尤为重要 平衡区面积不大或资源组合不是特别灵活的情况下，即便风电比重低于10%，预测的作用也非常关键 尽管风电预测还不能尽善尽美，当前的风电预测技术对电网运营商也是非常有帮助的 有希望节约大笔的机组组合和燃料消耗成本 未来提高风电预测水平的承诺 美国几家电网运营商正在进行风电预测 风电预测的趋势正在由分散式预测向集中式预测发展演变。 Exeter Associates, Inc.

Western Wind and Solar Integration Study (2010) 西部风电及太阳能并网研究(2010) “Using state-of-the-art wind and solar forecasts and unit commitment is essential and would reduce annual WECC operating costs by up to $5 billion ($4 billion in 2009 $) or $12-20/MWh ($10-17/MWh in 2009$) of renewable energy, compared to ignoring renewables in the unit commitment process.” “Perfect forecasts would reduce annual costs by another $500 million ($425 million in 2009 $) or $1-2/MWh ($0.9-1.7/MWh in 2009 $) of renewable energy, 运用当前最先进的风电和太阳能预测技术以及机组组合非常必要；与忽视可再生能源电力在机组组合里的作用相比，充分运用这些，可以为WECC每年减少多达50亿美元的运行成本（2009年是40亿），或者，帮助可再生能源减少一定的单位发电成本：每千度电可减少12到20美元的成本（2009年实际每千度电减少10到17美元的单位发电成本） “完美的预测还能把全年运行成本再减少5亿美元（2009年是4.25亿），或，帮助可再生能源再减少单位发电成本：每千度电可再减少1到2美元的成本（2009年再减少0.9到1.7美元） Exeter Associates, Inc.

Solutions: Develop More Flexible Electricity Markets 解决方案：建设更灵活的电力市场 Operational flexibility valuable to operators but of little value to power suppliers unless they are somehow compensated Market and policy changes will likely be necessary Expanding ancillary service markets Incentives for greater generator flexibility (existing and new plants) Changes to market requirements,(e.g., load following is not a compensated ancillary service) Make changes to accommodate wind ramps, which more closely resemble large load ramps More closely aligned with 10-30 minute non-spinning and supplemental reserves; current rules require these to be in-service for 1-2 hours when wind ramps can occur over several hours Forces operators to use more expensive regulation service instead Multiple ways to do this, depending on electricity system structure 灵活运行对电网运营商来讲很有价值，但是对电力供应商来讲却不是这样，除非可以在一定程度上给他们补偿 市场和政策变化很可能无法避免 扩展辅助服务市场 鼓励发电商加强灵活运行的激励机制 (新、老电厂) 调整市场要求/标准,(比如负荷跟踪不作为付费辅助服务) 做出调整以适应风电爬坡特性（很类似大型线性负荷） 与10-30分钟非旋转备用和补充电源更紧密地配合；现行规定是当风电爬坡可能持续几小时的情况可以开动使用这些备用设施1-2小时 要求运行商采用其他更昂贵的调节服务 实现目标的方法有很多，能采用哪些方法取决于电力系统的结构 Exeter Associates, Inc.

Solutions: Operate Over Larger Market Operation Areas 解决办法：扩大市场运行区域 There are about 140 balancing areas in the U.S. with wide variations in size, generation resources, and load Each must balance generation and load within its area Larger balancing areas have more access to generating resources to provide ancillary services Larger balancing areas can also take advantage of the geographic diversity of wind resources, helping to smooth wind variability ISOs and RTOs naturally capture the benefits of large balancing areas Smaller areas can gain some of those benefits through sharing agreements such as the ACE Diversity Interchange that pools area control error among utilities in the western U.S. 美国有140个左右的平衡区，大小、 电源结构和负荷情况各不相同 每个平衡区都必须平衡本区域内的发 电运行和负荷 大平衡区有更多的电源资源可供调配 以提供更好的辅助服务 大平衡区可以利用地理跨度大的优势， 削弱风资源变化的影响，平滑风电出 力波动 ISOs和RTOs 有大平衡区的先天优势 小平衡区也可以通过分享协议，如“ACE一体化交换”，扩展平衡区域，在美国西部的电力公司间更大范围地调度。 Exeter Associates, Inc.

如果应用小时内调度，风电波动的调节需求只有不到原来的三分之一 Less than 1/3 of regulation is necessary if sub-hourly scheduling is utilizied 如果应用小时内调度，风电波动的调节需求只有不到原来的三分之一 Exeter Associates, Inc.

Wind Integration Costs Lower in Larger Balancing Areas 大平衡区的风电并网成本更低 Date 时间 Study 研究单位 ISO/RTO Wind Level 风电占比 Integration Cost 并网成本($/MWh) Energy Market Interval 能源市场时间频度 March 2005 NYISO 10% Very Low 5 minute December 2006 Minnesota/MISO 31% $4.41 March 2007 Avista No 30% $8.84 1 hour Idaho Power $7.92

Summary 总结 截止2008年，美国的风电占整体电力供应的1-2%左右 很多电力公司新增大量的风电装机 Wind contributes about 1-2% of total U.S. electricity generation as of 2008 A number of utilities are adding significant amounts of wind capacity Over a dozen wind integration studies have been conducted to assess the technical ability and feasibility to incorporate wind energy In general, studies have progressed from “can it be done” to “how and at what cost” The studies have found that large interconnected power systems can accommodate high levels of variable energy by: Implementing a wind forecasting system Acquiring flexible generating resources Instituting new operating strategies for minimum load hours and other high risk periods; and Creating larger balancing areas, new market rules, and grid codes 截止2008年，美国的风电占整体电力供应的1-2%左右 很多电力公司新增大量的风电装机 开展了十来个风电并网研究，评估接入风电的技术能力和可行性 总的来讲，研究的内容已经从“能不能”发展到“怎么做、多大代价做” 研究表明，通过以下办法，大规模互联电网能够接纳较多的间歇式电源: 应用风电预测系统 利用灵活性电源 针对负荷低谷和其他高风险时段制定新的运行办法 创造大平衡区，新市场规则和并网规定 Exeter Associates, Inc.

Doing a Wind Integration Study 进行风电并网研究 Exeter Associates, Inc.

Typical Study Scope 典型的研究范围 Identify operational and cost issues from increasing levels of wind penetration Usually multiple scenarios of base, 10%, 20%, 30% Note: The study is about net wind impacts (load minus wind), not just wind alone, as the variations of load and wind partially cancel each other out Determine scope of study (all of China, part of China) If not modeling all of China, determine how to model areas external to study All assumptions, data inputs, and study results should be public to encourage maximum participation and overall support of the study results 找出风电比例提高后带来的运行和经济问题 通常按照风电整体电源占比10%, 20%, 30%的几种情景开展研究 注意: 研究这些问题的时候，关注的是风电的“净影响”（剔除了风电和负荷能互冲的部分），而不仅仅是风电本身的影响，因为负荷变化也可以抵消一部分风电的影响。 确定研究范围（整个中国，或部分地区） 如果不是针对整个中国研究，则需确定如何选取研究目标区域 所有的假设、数据和研究结果应该对外公布以实现广泛参与，促进研究结论得到认可和支持 Exeter Associates, Inc.

Typical Study Scope (2) 典型的研究范围（2） Identify changes in production costs (and cost volatility) from higher levels of wind penetration Identify changes in requirements for reserves over multiple time frames Consider whether existing generating resources (i.e., coal, natural gas, hydro) can accommodate higher levels of wind generation 风电比重加大后，发电成本的变化（成本波动） 不同时段对备用电源要求的变化 考虑现有电源结构（比如煤电、天然气发电、水电）是否能接纳更多的风电 Exeter Associates, Inc.

What a Wind Integration Study Does Not Do (Unless Designed to Do So) 风电并网研究可以忽略哪些问题 （除非设计意图如此） Additional maintenance and forced outages and derates Degradation of heat rate from ramping, cycling and range of operation 附加维护、被动停发和减发 爬坡、循环和运行范围引起的热效率下降 Exeter Associates, Inc.

Wind Integration Studies Generally Have Multiple Parts 风电并网研究通常包含几个部分 Production cost modeling (for hourly simulations of power flows, costs and dispatch) Statistical analysis (for determining hourly deviations in net load from adding wind, and for determining sub-hourly changes) Load flow analysis Transmission planning and siting 发电成本模型（潮流、成本和调度的小时间隔模拟） 统计分析(确定因增加风电导致的净负荷每小时偏离，及小时内变化) 潮流分析 输电线路规划和选择 Exeter Associates, Inc.

Organizational Set-Up 分工 Study time-consuming and labor-intensive Typical study time 1 year, and can take longer Need single organization to lead and take responsibility With help from other organizations and from Technical Review Committee Consider who does the work Could be done in-house but hard to juggle wind integration study with other responsibilities Work typically is outsourced to multiple consultants Power systems consultant Transmission/load flows consultant Wind resource and forecasting consultant 研究所需时间和人员付出 一般来讲需要一年时间，可能还要更长 需要一家机构牵头和负责 同时需要其他机构和技术审查委员会的帮助 考虑由谁来做 可以内部来做，但是很难兼顾风电并网研究以外的其他问题和事情 通常把一些工作外包给多个咨询专家/公司 电力系统专家 输变电/潮流专家 风资源和风电预测专家 Exeter Associates, Inc.

Technical Review Committee 技术审查委员会 Comprised of grid operators, government energy officials, renewable energy companies, renewable energy experts and/or consultants Help design study objectives and approach and act as peer reviewers Consider quarterly meetings, both in-person and by telephone Provide education to all committee participants 包含电网运营商、能源部门的政府官员、可再生能源发电公司、可再生能源专家和顾问 帮助设计研究目标和方案，并帮助评议 考虑召开季度会议，既可以当面开会交流，也可以电话会议 为成员提供培训 Exeter Associates, Inc.

Scenarios 情景 Include multiple scenarios of increasing wind penetration, by energy Start with base year (i.e., current year or the last year with most complete set of data) Include at least one scenario with very high wind penetration (e.g., 30% or 40%) Consider scenarios focused on particular circumstances, e.g., geographically diversified wind, high offshore wind, high transmission, etc. 风电在能源结构中（依据发电量）占比递增的多种不同情景 起始基年(比如有全套数据的当年或上一年) 至少包含风电比例很高的一种情景(比如30% 或 40%) 着重考虑特殊情况下的一些情景, 如地理分布较为分散的风电, 有较多的海上风电,较大的送电需求等 Exeter Associates, Inc.

Scenarios 情景 风电在能源结构中（依据发电量）占比递增的多种不同情景 起始基年(比如有全套数据的当年或上一年) 至少包含风电比率很高的一种情景(比如30% 或 40%) 着重考虑特殊情况下的一些情景, 如地理分布较为分散的风电, 有较多的海上风电,较大的送电需求等 Include multiple scenarios of increasing wind penetration, by energy Start with base year (i.e., current year or the last year with most complete set of data) Include at least one scenario with very high wind penetration (e.g., 30% or 40%) Consider scenarios focused on particular circumstances, e.g., geographically diversified wind, high offshore wind, high transmission, etc. Exeter Associates, Inc.

Scenarios (2) 情景（2） Escalate generation (including wind) and load data by annual constant percentage to get to future year (e.g., 2%) Don’t project too far out into the future, as it may be difficult to control all the potential variables Studies typically forecast out 10 years May need to add generation over study projection years to maintain reliability U.S. studies typically add gas turbines as proxy 基于年度固定百分比（如2%）逐步升高的发电量 (包括风电)和负荷数据 不要把预测的期限定的太远, 因为很多潜在的变量是难以控制的 一般来说研究预测的期限是10年左右 为保持可靠性，可能需要在预测研究的基础上再增加发电量 美国这边的研究通常以天然气发电量的增长来代表 Exeter Associates, Inc.

Sensitivity Studies 敏感度研究 May wish to do several sensitivity studies Different natural gas and coal prices Different fuel mixes Higher or lower projected electricity demand More demand response Plug-in electric vehicles Different turnback capabilities for coal units Differing levels of hydro flexibility Varying accuracy levels of wind power forecast 可能要做几种敏感度研究 不同的天然气和煤炭价格 不同的燃料组合 高一点或低一点的电力需求预测 更好的需求响应 插入式电动车 不同的燃煤机组爬坡控制能力 不同水平的水电灵活性 风电预测的不同准确度 Exeter Associates, Inc.

Assumptions to Make 假设/预测 Projected future fuel costs for natural gas and coal Expected future load growth Projected future transmission grid Projected generating capacity additions Whether to include $/ton carbon dioxide adder or not 预测未来天然气和煤炭的燃料价格 预期未来的负荷增长 预测未来的输电网络 预测发电装机的增加 是否包含二氧化碳减排成本（美元/吨） Exeter Associates, Inc.

Data Needs 数据需求 Multiple years (usually three) of time-synchronized wind and load data to capture inter-annual variability of load and wind Use Numerical Weather Prediction models to recreate historical weather data 多年（一般三年）的同期风电及负荷数据，用以推测风电和负荷情况的年际变化 使用数值模拟气象预测模型重现历史气象数据 Exeter Associates, Inc.

Examples of Wind Data 风电数据示例 Wind speed data extracted from multiple-year model runs that are at wind turbine hub heights and converted to wind power using wind power curves Output is typically 5-minute or 10-minute wind data Tons of data Use actual wind generation to verify modeled wind data Need to factor in land use restrictions such as urban areas, national parks, environmentally sensitive areas, and other areas unlikely to be developed 从多年模型中取得轮毂高度风速数据，并利用风功率曲线换算为风能 一般来讲，输出的风资源数据格式是5分钟或10分钟间隔 海量数据 使用实际风电发电量来核实模拟的风数据 需要考虑土地征用限制如城市范围、国家公园、环境敏感区、及其他相关因素，排除不能开发风电的地区。 Exeter Associates, Inc.

Data Needs (2) 数据需要（2） Sub-hourly load and generation data (minute-by-minute, 10-minute) for analysis of interesting periods (high wind, high load, high wind/low load, etc.) Load forecasts for multiple years This step always takes the longest, and the integration study cannot proceed without data 通过逐时负荷和发电数据（每分钟间隔、10分钟间隔）做重点时段分析（高比例风电出力，高负荷，高风电/低负荷，等） 多年负荷预测 这步工作用时最长，但并网研究缺了这些数据就没法继续 Exeter Associates, Inc.

Wind Forecasts 风电预测 Techniques for simulating wind power forecasts not well developed Two basic methods Impose random error on top of wind generation profiles Match up with another wind forecast (covering same time frame) 风电预测模拟技术尚未完全成熟 两个基本方法 对风力发电数据加以随机误差分析 与其他风电预测比对 (相同时段) Exeter Associates, Inc.

Transmission 输电 Need current and projected load flows for planned scenarios Also need to decide whether to do detailed transmission study for low-voltage and high-voltage transmission levels, or do high-voltage schematic and assume low-voltage issues away 确定输电预期情景，需要当前和预估的潮流情况 也需要决定是否针对低压输电和高压输电做详细的输电研究，或（假设不存在低压问题）只做高压方案研究。 Exeter Associates, Inc.

Wind Integration Costs 风电并网成本 In general, defined as the increased system costs imposed by higher levels of wind generation, but different viewpoints on how to calculate wind integration costs Two common methods: Flat block, where a power system with wind is compared to a system with an energy-equivalent flat block, and the cost difference is the integration cost System, focusing on wind’s impact on commitment and dispatch and wind’s net value from fuel savings and reduced wholesale prices Newer, but still relatively untested, is using an “ideal wind block” (perfectly forecasted) instead of a flat block 风电并网成本，一般来说，主要都是指因风电比重加大而导致的系统成本上升，但在如何计算方面，存在一些不同看法： 两种一般计算方法: “水平出力”（假设发电输出为恒定），比较并入风电的电力系统和相应能当量无风电的电力系统的水平出力，差额就是并网成本 主要看风电 接入对电力系统机组组合、调度的影响，以及风电节约燃料贡献的净价值、降低的电力批发价格 有种新的方法，但相对欠缺验证，就是用“理想风电 出力”(基于完美预测)取代“水平出力” Exeter Associates, Inc.

Study 研究 Inputs, assumptions, data should be public and transparent Study results should also be made public Non-disclosure agreements may be required to get access to sensitive data 内容、假设、数据等应该公开透明 研究结果也要公布 对于敏感数据可能需要签订保密协议 Exeter Associates, Inc.

Questions to Consider 需要考虑的问题 Would a methods workshop in December be valuable? Would it make sense to do a wind integration study in China? If yes, how to begin? 12月召开研究方法的研讨会是否很有意义？ 在中国开展风电并网研究有意义么？ 有意义的话，如何开始？ Exeter Associates, Inc.

Contact Information 联系方式 Kevin Porter Exeter Associates, Inc. 10480 Little Patuxent Parkway, Suite 300 Columbia, MD 21044 United States of America 1+410-992-7500 1+410-992-3445 fax kporter@exeterassociates.com