Alibaba Cloud Energy and Utilities China Translated from Chinese

China Southern Power Grid (南方电网)

Real-time Analytics · Risk Management

China Southern Power Grid needed a more elastic and data-rich dispatch architecture to manage fast-growing market complexity, new energy sources, and weather-sensitive grid operations. Its traditional centralized systems were hard to scale, hard to upgrade, and unable to turn years of protected dispatch data into reliable AI and forecasting capabilities.

7-day 9 km typhoon forecast dropped from 8 hours to 1.7 hours
Shenlong platform reaches 825 trillion floating-point operations per second
capacity expansion takes 10 seconds
full system one-click deployment takes 70 seconds

Reported impact

Business outcome

The utility raised dispatch accuracy, shortened deployment and scaling cycles, and created a cloud-HPC base for weather and market decisions that directly affect grid reliability and cost. The same platform now supports AI load forecasting, faster market settlement, and more accurate extreme-weather response across a supply area serving hundreds of millions of people.

  • Cost Reduction
  • Customer Experience
  • Productivity Improvement
  • Scalability

Solution

China Southern Power Grid and Alibaba Cloud transformed the utility's dispatch platform from a concentrated architecture into a distributed cloud foundation and then added a supercomputing layer for AI- and weather-intensive workloads. The dispatch cloud uses multi-region design, unified identity and monitoring, hierarchical operations, and distributed stream, batch, and AI processing so grid data can be modeled instead of sitting idle on disk. On top of that base, the utility built a cloud-native SCADA platform that can expand compute resources in seconds, process more than sixty thousand messages per second, and deploy the whole system in about seventy seconds. In 2021 it also launched the power sector's first dispatch-cloud supercomputing platform on Apsara and Shenlong infrastructure, which gave the grid elastic high-performance compute for fine-grained weather prediction, typhoon response, spot-market clearing, and new-energy power forecasting. The combined platform turns dispatch, forecasting, and market operations into one data-driven AI control plane.

Data flow

Dispatch, market, and operational telemetry is ingested into the distributed dispatch cloud, where streaming, batch, and AI services analyze and model grid behavior. The results feed cloud-native SCADA for real-time operations and the supercomputing layer for heavy tasks such as weather prediction, power-market clearing, and new-energy forecasting before dispatch teams act on the outputs.

Solution architecture

4 components · 3 layers
  1. Compute
    • Shenlong Supercomputing Cluster Supplies the elastic high-performance compute used for weather forecasting and market calculation.
    • Realtime Compute for Apache Flink Processes streaming grid data as part of the distributed dispatch architecture.
  2. Serving
    • Cloud-native SCADA Handles second-level elastic dispatch control and high-throughput operational messaging.
  3. Orchestration
    • Dispatch Cloud Platform Provides the distributed dispatch foundation with unified identity, monitoring, and distributed data processing.

Architecture clues

  • SCADA is rebuilt as a cloud-native service with second-level elasticity and extremely high message throughput.
  • The dispatch redesign combines distributed stream, batch, and AI processing instead of adding AI to an unchanged centralized control plane.
  • The supercomputing layer is exposed as cloud services so downstream applications do not need to be rewritten to consume HPC capacity.

Evidence from the source

On the distributed dispatch platform, South Grid built the first cloud-native SCADA system in the industry with 60,000+ messages per second, 70-second one-click deployment, and 10-second expansion. (original: 研发出业内首个可秒级扩展计算资源、即插即用的云化 SCADA 系统... 实现了每秒处理消息超过 6 万条、全系统 70 秒一键部署、10 秒快速扩容。)
The dispatch cloud raised whole-network load-forecast accuracy to 97.63%, and in 2021 the company launched the power sector's first dispatch-cloud supercomputing platform. (original: 基于调度云平台开发的 AI 应用,将南网总调的全网负荷预测准确率提升到 97.63%。2021 年,南方电网与阿里云合作上线国内电力行业首个调度云超算平台。)
The supercomputing platform can perform 825 trillion floating-point operations per second, cutting 7-day 9 km typhoon forecasting from 8 hours to 1.7 hours. (original: 每秒钟可进行 825 万亿次浮点运算... 对于未来 7 天台风 9 公里的预测... 只需用 16 台超算服务器运行 1.7 个小时。)
Transparency Evidence & methodology 100/100 reference value · 88% confidence · 1 source

Review how this reference was scored and trace each normalized claim back to its documented source.

01

Scoring methodology

How scores are calculated

Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.

Customer reference value100/100
  • Problem clarity20/20
    Measures whether the customer's operational problem is specific enough to compare.
  • Implementation depth25/25
    Rewards named components, data flow, integrations, and architecture detail.
  • Outcome specificity20/20
    Measures whether the reported benefit is concrete rather than a generic claim.
  • Quantified impact25/25
    6 metrics describe measurable impact.
  • Deployment scale10/10
    Rewards evidence of a live or scaled deployment and meaningful integration scope.
Evidence confidence88%
  • Source authority20/25
    The main evidence is an official vendor-published customer story.
  • Claim traceability25/25
    12 of 12 key fields link to a named source.
  • Direct evidence15/20
    3 direct source quotes retained.
  • Date provenance10/10
    The publication date comes from the source page or publisher.
  • Source transparency8/10
    1 source reference document publisher, URL, authority, retrieval, and supported fields.
  • Implementation specificity10/10
    Rewards source-backed components, data flow, quantified impact, and a specific outcome.
02

Traceability

Sources & provenance

Field coverage shows which facts each source supports. Dates are shown only when documented by the source.

  1. Alibaba Cloud

    南方电网:智能化的“最高调度指挥官”_电力-阿里云客户案例

    First-party
    https://www.aliyun.com/customer-stories/power-industry-2022-csg
    Source type
    Vendor-published customer story
    Supports
    • Customer
    • Region
    • Company size
    • Business problem
    • Solution
    • Products
    • Solution components
    • Data flow
    • Integrations
    • technical area
    • use case category
    • Business outcome
    • outcome category
    • Metrics
    • Architecture
    • evidence quotes

Date evidence

  • Published on source page Dec 20, 2023 · Alibaba Cloud · First-party
    Open date source
Not separately sourced Industry. Treat these normalized attributes as lower-confidence until a dedicated source is added.