Valuation and Integration of Solar Energy: 太阳能评价与并网 Early Results from the United States 美国初期研究成果 Ryan Wiser 怀瑞恩 Consultant to the Center for Resource Solutions 资源解答中心专家 Energy Foundation Workshop 能源基金会研讨会 Beijing, China July 2, 1010

Motivation 研究动机 Increasing focus on solar as source of electricity: Growth in interest in large-scale plants: >50 GW seeking interconnection in U.S., ~10 GW in later stages of development Decision-makers need to understand electric system value of solar electricity (PV/CSP, w/ and w/o storage): Solar is variable/unpredictable: valuation will differ from conventional gen. Output variability/uncertainty studied for wind; little similar work on solar Two areas of particular interest: Estimating the long-term electric system value of solar without storage given displacement in conventional gen. and optimal capacity expansion Understanding the technical and economic impacts of short-term variability and uncertainty in PV output, and associated integration costs Current presentation does not include power quality timeframe 太阳能发电受到越来越多的关注 美国对大规模电站的兴趣在提升：预期的并网需求超过50GW，约10GW的项目处于开发的后期阶段 决策者需要认识到太阳能电力在电力系统中的价值（光伏/光热，有无储能） 太阳能具有波动性/不可预测性：其评价方法将有别于常规能源 针对风电出力变化/不确定性已有研究；太阳能这方面的研究却很少 两个主要研究方向： 估算无储能太阳能在电力系统中替代常规电力的长期价值，及最佳容量 了解光伏发电输出短期波动性和不确定性带来的技术和经济影响及并网成本 目前的报告并不包括瞬时电能质量

Long Term Valuation 长期价值

Value of an Electricity Generation Source Impacted By Multiple Factors 电源价值受多种因素影响 发电成本 输电成本 并网成本 容量价值 能源价值

Central Solar in Western U. S Central Solar in Western U.S. Currently >3 cents/kWh More Valuable Than Wind Considering Energy/Capacity Value 考虑到能量/容量价值，美国西部集中式太阳能价值比风电高3美分/度 Increased value due to coincidence with system peak, and therefore offset of higher-cost resources; benefits likely to be lower in China 因与用电负荷高峰重合而增加了太阳能电力的价值，所以抵消了部分高成本；在中国，这方面的价值可能要低一些

Marginal Value of Solar Energy Drops with Penetration, But At Current Levels, Added Value Is Considerable 随着在总体电源中比重的增加，太阳能的边际价值会减少，但在目前的水平下，增值仍相当可观 Solar reduces peak load (high capacity value) Solar shifts peak load into night hours (low/no capacity value) Solar increasingly curtailed during peak production Renewable Penetration (% energy basis) 可再生能源比重（发电量）

For PV (less so for CSP), Lower Transmission Needs Further Augment Added Market Value 对于光伏（光热的情况不太一样），较低的输电需求进一步增加了它的市场附加值 Figure shows estimated transmission cost for wind in US for large number of studies: > 1 cent/kWh 基于大量研究，数据表明美国风电的输电成本高于1美分/度电 Solar PV is modular and does not experience as strong economies of scale, allowing siting closer to load in many cases 太阳能光伏是由一个个独立模板组成的，规模经济效益不如其他电源明显，并且很多情况下可以建在负荷周边

But, Solar without Storage Is Also Variable and, to a Degree, Unpredictable 但是，无储能太阳能仍然是波动性较大的，某种程度上，不可预测 Two key assumptions in previous analysis Perfect foresight (did not evaluate forecast error) No sub-hourly variability (changes within hour require balancing) Costs for dealing with imperfect forecasts / sub-hourly variability = “integration costs” Wind integration well studied: cost < $10/MWh Integration costs for solar poorly understood Note: Focus here is on large-scale PV; CSP with and without storage a much smaller concern due to thermal inertia and storage 之前分析的两个关键假设 完美预测（预测误差未评估） 不存在小时内波动 (需要平衡小 时内波动) 预测误差带来的成本/ 小时内波 动= “并网成本” 大量研究表明，风电并网成本 < $10/MWh （10美元/千度电） 但对太阳能并网成本认知还太少 注意: 此处关注的是大规模光伏; 有 储能或无储能的光热因其热惯量和 储存问题，受关注更少

Short Term Integration 短期并网

Clouds Can Produce Rapid Ramps in PV Output 阴云会导致光伏出力锐减 PV plant output can change by 70% in five minutes at single site 一个光伏电站的发电出力，在五分钟内变化可达70% Actual PV plant output on a clear and cloudy day for a plant in the Southwest United States

Operational Impacts Vary with Time Scale 针对不同时段运行影响有所不同 Regulation: Automatic adjustments to conventional generation output to maintain balance between supply and demand 调节：自动调整至常规发电输出保持供需平衡 Load Following: Periodic operator dispatch of generation to follow trends in load 负荷跟踪：分时段运行调度以配合负荷变化趋势 Scheduling: Commitment of generation units based on forecasted need for generation 调度计划：基于需求预测的发电约定 Source: Milligan 2009

Concerns that Rapid Fluctuations in Output Are a Potential Roadblock to PV 对光伏出力快速波动的担忧有可能阻碍光伏的发展 Market operators / utilities have expressed concern 市场运营商/电力公司表示这样的担忧 Numerous early studies found potential limits to increased PV penetration 大量早期研究给出了光伏电源整体占比的可能极限 But… many of early concerns did not account for geographic diversity and forecasting 然而 …许多的担忧并未基于全面考虑，分散的地理分布因素和预测贡献的因素都未考虑 Solar PV’s characteristics allow for ready use of site diversity and forecasting 太阳能光伏的特征允许分散式布局和利用预测技术

太阳能电站有波动性，但分散的地理分布可以平滑这种出力波动 Solar Plants Are Variable, But Output Smoothing from Geographic Diversity 太阳能电站有波动性，但分散的地理分布可以平滑这种出力波动 Clouds cause rapid changes in output of individual plants 就单个电站而言，阴云会引发快速的出力波动 Multi-MW plants can change by > 70% in less than 5-min 兆瓦级电站在五分钟内变化可达70%以上 Aggregation dramatically reduces variability; even within-plant variability at < 1 min reduced with plant size due to smoothing 将众多电站作为整体考虑，则波动会大幅降低；同理，即便是单个电站范围内，一分钟内波动也随电站规模扩大而减少 Source: Wiemken 2001 Actual 5-min data from 100 PV plants in Germany. Aggregate output (P) is normalized by installed capacity (P installed)

整体优势在电站分布不是非常集中的情况下可以降低相对波动性 Aggregation Decreases Relative Variability When Sites Are Less Than Perfectly Correlated 整体优势在电站分布不是非常集中的情况下可以降低相对波动性 Deltas: step changes from one averaging interval to the next Clear Sky Index: ratio of insulation if no clouds were present to measured insolation Data: 1-min insolation data from 23 time-synchronized sites in the U.S. Source: Mills and Wiser 2009

阴云导致预测偏差，多云/晴转阴的天气会加大误差。不同地区多电站形成的整体优势可以降低预测误差。 Forecasting for Solar Still Developing, But Is Much Better for Geographically Diverse Sites 太阳能预测技术仍在发展，针对地理上分散分布的电站来讲预测效果更佳 Clouds lead to forecast errors, and errors are greatest on partly cloudy days Forecast errors for geographically diverse sites smoothed by aggregation 阴云导致预测偏差，多云/晴转阴的天气会加大误差。不同地区多电站形成的整体优势可以降低预测误差。 Source: Lorenz 2009 Forecasts of solar output based on forecasts from the European Centre for Medium-Range Weather Forecasts

Longer Forecast Horizon Increases Error 远期预测误差增大 Assuming that current cloud conditions will not change (persistence forecast) is reasonable for short forecast horizons (<1 h) Satellite or numerical weather model based forecasts are better for longer forecast horizons 对短期预报来说，假设当前云的条件不变(持续性预报) 是合理的（一小时内) 卫星或数值天气模型预报，对长期预测更为合适。

Multiple Sources of Reserves Can Mitigate Variability 多种类型备用电源可以减轻出力波动 Description Timescale Mechanical inertia Rotating mass of generator Instantaneous, sustain for 10s Thermal inertia Steam pressure in boiler 10s, sustain for mins Storage/hydro Turbines not currently used 10s to 10 min Gas turbines Peaking and backup turbines 5 min to 20 min Thermal plant overrunning Run station above max rated capacity Can run 3-10% over for few min without added stress Industrial load management Voluntary interruptable loads Fast response but need to make provisions ahead of time Domestic load management Automatic interruption of some loads (i.e. water heating) Instantaneous with proper equipment installed Voltage reduction Low voltage reduces loads Instantaneous, but degrades supply Imported power Additional flow on tie lines Instantaneous, need to arrange/warn Spinning reserve Spare capacity on committed units Large thermal units can provide 2-4%/min reserve up to capacity Banked/pre-scheduled plant Units on hot standby or scheduled early Available in 1-2 hours Source: Grubb 1991

Additional Reserves Have Modest Cost for Diverse Solar Sites 太阳能电站分散分布可以大大降低备用电源成本 Source: Mills and Wiser 2010 5 close sites: ~ 7,000 sq. km 25 site grid: 5 X 5 Site array with 40 km spacing between sites ~ 40,000 sq. km These costs do not include cost of day-ahead forecast error. EnerNex (2009) estimated day ahead forecast costs to be $4-7/MWh for 800 MW PV/CSP combination, depending on assumptions for fuel costs 这些成本不包含一日前预测误差。EnerNex (2009) 估算一日前预测成本，800兆瓦的光伏/光热组合规模下，大概是4至7美元每千度电， 这个成本也取决于对燃料成本的预期。

Conclusions from the United States 美国经验总结 Solar energy without storage, at low levels of penetration, can be quite valuable within the US energy system, supporting higher tariffs and policy measures Peak coincidence increases energy value High resource adequacy / capacity value Lower transmission costs compared to less-modular technologies Implications for China less clear… Short term variability and uncertainty in solar without storage creates integration issues that appear to be relatively similar to wind in terms of cost implications Given rapid pace of technology and cost developments, important to start thinking about large-scale solar valuation and integration at an early stage 太阳能在整体电源中占比较低情景下，无储能太阳能发电在美国电力系统中较有价值，值得给以高电价和政策支持 与负荷高峰的重叠增加其能源价值 资源量充足/ 容量价值高 输电成本相对低 对中国的借鉴意义尚不明晰 无储能太阳能的短期出力波动性和不确定性带来相关的并网问题，增加了成本，这点和风电相似 鉴于技术进步和成本下降的速度较快，有必要提前考虑大规模太阳能发电的评价和并网问题