The 100 Largest Hydropower Plants in the Greater Mekong Subregion

China plays an increasingly large role in global energy and electricity sector development through its overseas investment. The Chinese hydropower industry has come to dominate global hydropower development, particularly in the Greater Mekong Subregion (GMS), which consists of the countries along the mainstream Mekong River – Laos, Myanmar, Vietnam, Thailand, and Cambodia.

In order to investigate patterns of Chinese and other overseas hydropower developers in the GMS region, we assembled a dataset of hydropower projects in the five GMS countries excluding China. This unit-level dataset was initialized by merging four datasets: the Platts World Electric Power Plants database, the Stimson Center Mekong Infrastructure Tracker, Boston University China’s Global Power Database, and the WRI Global Power Plant Database. Each of these databases has varying technological and financial information at the plant level.

We first aggregated the Platts database by plant rather than unit for uniformity. Then we used a Ruby script to build a concordance table, which matched plant names used in different databases. The script iterates through all databases except WRI’s GPPD, applying each of four strategies to the plant name string to find matches across data sources: 1) manual overrides, 2) naive string match, 3) ignoring suffixes, 4) using alternates. Next, the script iterates through the unmatched plant names and aggregates plants by phase. If one source has a project aggregated by phase where other sources list each phase separately, the phases are added to the concordance table to match the aggregated project name. Then the WRI GPPD is incorporated into the concordance table by matching on Platts database ID or plant name where no ID is available. Finally, a merged database including all attributes across data sources was constructed using the concordance table, where capacity is summed when aggregation is required.

We selected the 100 largest hydropower plants, in terms of capacity, for verification and collection of additional data. When not already identified by the Platts data, we performed a web search to identify the company name and country of origin for developer companies, their parent companies, and construction companies for each plant. Where there was a discrepancy between the four datasets on basic characteristics of the plant – namely, capacity, status, or year, we performed a web search to find a reliable and recent source of information to confirm the data. We also used a literature review and web search to fill in data on key features related to the size of the associated dam and reservoir for each plant. For discrepancies and missing data, we welcome feedback and contributions from users of this dataset.

中国在全球能源与电力行业的发展中扮演着日益重要的角色，其海外投资对此起到了推动作用。中国的水电产业已逐渐主导全球水电开发，尤其在湄公河次区域（GMS）——由湄公河主流流域国家老挝、缅甸、越南、泰国和柬埔寨组成——的表现尤为显著。

为了探究GMS区域内中国及其他海外水电开发商的运作模式，我们搜集了一个涵盖五个GMS国家（除中国外）水电项目的单元级数据集。该数据集由四个数据源合并而成：普拉茨全球电力发电厂数据库、斯廷森中心湄公河基础设施追踪器、波士顿大学中国全球电力数据库以及世界资源研究所全球电力发电厂数据库。这些数据库在发电厂层面提供了不同技术及财务信息。

我们首先对普拉茨数据库进行了按发电厂而非单元的汇总，以实现数据的一致性。随后，我们使用Ruby脚本构建了一个对照表，该表匹配了不同数据库中使用的发电厂名称。脚本遍历所有数据库（除世界资源研究所的全球电力发电厂数据库外），对发电厂名称字符串应用四种策略以在数据源间寻找匹配：1）手动覆盖，2）简单的字符串匹配，3）忽略后缀，4）使用替代名称。接着，脚本遍历未匹配的发电厂名称，按阶段汇总发电厂。如果一个来源按阶段汇总项目，而其他来源则单独列出每个阶段，则将这些阶段添加到对照表中以匹配汇总的项目名称。然后，通过匹配普拉茨数据库ID或发电厂名称（如果不存在ID），将世界资源研究所的全球电力发电厂数据库纳入对照表。最后，利用对照表构建了一个包含所有数据源属性的综合数据库，在需要汇总时将容量相加。

我们选取了100个按容量计算的最大的水电厂进行验证和数据收集。当普拉茨数据未识别出开发公司、母公司以及每个发电厂的建设公司名称和国籍时，我们进行了网络搜索以识别这些信息。在四个数据集关于发电厂基本特征（即容量、状态或年份）存在差异时，我们进行了网络搜索，以找到可靠且最新的信息来源以确认数据。我们还通过文献综述和网络搜索补充了与每个发电厂相关的坝体和水库关键特征的数据。对于数据中的差异和缺失信息，我们欢迎数据集使用者提供反馈和贡献。