SCADA系统
一、SCADA架构 | SCADA Architecture
二、RTU与现场设备 | RTU & Field Devices
三、SCADA通信 | SCADA Communication
四、SCADA应用场景 | SCADA Applications
Solar Farm SCADA Architecture Design — 200MW PV Plant
Alright team, we need to finalize the SCADA architecture for the 200MW Desert Sun Solar Farm. This plant has 400,000 PV modules, 2,000 string inverters, 80 inverter stations, a 230kV substation, and two 100MWh BESS containers. We need full visibility and control from the central control room and remote access from the corporate office 300 km away. Li Wei, walk us through the proposed architecture.
好了团队，我们要敲定200MW沙漠光伏电站的SCADA架构。这个电站有 40万块光伏组件、2000台组串式逆变器、80个逆变器方阵、230kV升压站、 两套100MWh储能集装箱。我们需要从中央控制室完全监控、并从300公里 外的公司总部远程访问。李伟，讲讲建议的架构。
(projecting architecture diagram) We're proposing a three-tier architecture. Tier 1 — field level: the 2,000 string inverters communicate via RS-485 to 80 data loggers in the inverter stations. Also at tier 1, the substation IEDs — protection relays, meters — use IEC 61850. The BESS and meteo stations use Modbus TCP.
（投影架构图）建议三层架构。第一层——现场级：2000台组串式逆变器 通过RS-485通讯到80个逆变器方阵的数据采集器。第一层还有升压站IED—— 保护继电器、表计——用IEC 61850。储能和气象站用Modbus TCP。
Tier 2 — plant control level: two redundant SCADA servers in the central control room, running a real-time database. They poll all 80 data loggers via Modbus TCP over fiber optic ring. The polling interval is 1 second for critical data — inverter output, grid frequency — and 5 seconds for non-critical data like module temperatures.
第二层——电站控制级：中央控制室两台冗余SCADA服务器、运行实时 数据库。通过光纤环网Modbus TCP轮询80个数据采集器。关键数据轮询 间隔1秒——逆变器输出、电网频率——非关键数据比如组件温度5秒。
1-second polling for 2,000 inverters — that's a lot of data. What's the bandwidth requirement on that fiber ring?
2000台逆变器1秒轮询——数据量不小。光纤环网带宽需求多大？
Each inverter station sends about 200 data points. With Modbus TCP overhead, that's roughly 5 KB per station per poll. 80 stations × 5 KB × 1 poll/second = 400 KB/s — about 3.2 Mbps. Our fiber ring is 1 Gbps. Even with 3× headroom for SOE events and BMS data, we're using less than 1% of the available bandwidth. Plenty of room.
每个逆变器方阵发送约200个数据点。加上Modbus TCP开销，大约每 方阵每轮询5 KB。80方阵×5 KB×1次/秒=400 KB/s——约3.2 Mbps。我们 光纤环网1 Gbps。即使考虑SOE事件和BMS数据3倍余量，也只用不到1% 可用带宽。空间大得很。
What about the grid interconnection? The utility requires us to send real-time data to their control center via DNP3 — they need to remotely curtail our output if there's a grid emergency.
电网互联呢？电力公司要求我们用DNP3实时送数据到他们调控中心—— 电网紧急时他们需要远程限电。
The SCADA server will act as a DNP3 outstation to the utility's master station. We'll map about 50 points — active power, reactive power, voltage, frequency, breaker status, and the curtailment control point. The DNP3 link is over a dedicated 4G cellular router with VPN — the utility doesn't allow internet-based connections.
SCADA服务器将作为DNP3子站对电力公司主站。我们映射大约50个点—— 有功功率、无功功率、电压、频率、断路器状态和限电控制点。DNP3链路 通过专用4G蜂窝路由器加VPN——电力公司不允许互联网连接。
And the remote access from the corporate office?
公司总部的远程访问呢？
We'll set up a read-only web client — the SCADA server publishes dashboards over HTTPS. The corporate users log in through a VPN with two-factor authentication. They can view all data but can't send control commands. For control, only the central control room operators have authority, and they must be physically present — control commands use hardware key switches on the operator console, not just software permissions.
我们设一个只读Web客户端——SCADA服务器通过HTTPS发布仪表盘页面。 公司用户通过VPN加双因子认证登录。可以查看所有数据但不能发控制 命令。控制只有中央控制室操作员有权限，而且必须物理在场——控制 命令用操作台上的硬件钥匙开关，不仅仅是软件权限。
Good. What about cybersecurity? The utility is very strict about NERC CIP compliance.
好。网络安全呢？电力公司对NERC CIP合规要求很严格。
We're following NERC CIP standards. The SCADA network is completely isolated from the corporate IT network by a firewall with a DMZ. All ports are locked down — only explicitly required ports are open. We have intrusion detection, antivirus on the servers, and quarterly vulnerability scanning. All user access is logged and auditable. And no wireless access points on the SCADA network — hardwired only.
我们遵循NERC CIP标准。SCADA网络通过防火墙加DMZ与公司IT网络 完全隔离。所有端口锁定——只开放明确需要的端口。我们有入侵检测、 服务器防病毒、季度漏洞扫描。所有用户访问都有日志可审计。还有， SCADA网络上没有无线接入点——全有线。
Pipeline Leak Detection — SCADA Alarm Response
(staring at the SCADA wall display) Sophie, I'm getting a leak detection alarm on the Northern Crude Pipeline — section KP-127 to KP-145. The SCADA leak detection system is showing a mass imbalance of 0.8% over the last 15 minutes. It's flagged as a "Level 2 — Possible Leak." But the pressure and flow trends look normal at the endpoints. Is this a real leak or a false alarm?
（盯着SCADA大屏）Sophie，北部原油管道收到泄漏检测报警——KP-127 至KP-145段。SCADA泄漏检测系统显示过去15分钟质量不平衡0.8%。标注 为"2级——可能泄漏"。但两端压力和流量趋势看着正常。这是真泄漏还是 误报？
Let me apply the diagnostic tools. (opens the leak detection module on her workstation) First, let's rule out instrumentation errors. The inlet flowmeter at KP-127 and the outlet flowmeter at KP-145 — both are ultrasonic. Let me check their diagnostic status... both show "healthy," no fault codes. Signal quality is above 95% for both. So the flow measurements are reliable.
我用诊断工具。（在工作站打开泄漏检测模块）首先，排除仪表误差。 KP-127进口流量计和KP-145出口流量计——都是超声波。我查诊断状态…… 两台都显示"健康"，无故障码。信号质量都95%以上。所以流量测量是 可靠的。
OK, so the 0.8% imbalance is real. But 0.8% of 5,000 m³/h is 40 m³/h — that's a significant leak if it's real. Over 15 minutes, that's 10 cubic meters of crude oil.
好，那0.8%不平衡是真的。但5000 m³/h的0.8%是40 m³/h——如果是真 泄漏那不小。15分钟就是10方原油。
Let me run the wave propagation analysis. (clicks through options) This system uses negative pressure wave detection — when a leak occurs, it creates a pressure drop that propagates in both directions at the speed of sound in the fluid. By measuring the time difference when that wave arrives at the upstream and downstream pressure sensors, we can calculate the leak location.
我跑一下波传播分析。（点击选项）这个系统用负压波检测——发生 泄漏时，产生的压力降以流体中声速向两个方向传播。通过测量该波到达 上游和下游压力传感器的时间差，能算泄漏位置。
(system processes for 30 seconds) The analysis found a negative pressure wave event at KP-138.2 — arrival time difference was 0.34 seconds. Estimated leak size: 35-45 m³/h. Confidence: 87%. This is NOT a false alarm, Mike. We have a real leak at approximately kilometer post 138.
（系统处理30秒）分析在KP-138.2发现负压波事件——到达时间差0.34秒。 估算泄漏大小：35-45 m³/h。置信度：87%。这不是误报，Mike。我们在 大约138公里标处有真泄漏。
(grabbing the radio) Jorge, this is Control. We've confirmed a leak on the Northern Crude line, approximately KP-138. Dispatch your patrol team now. I'm initiating the emergency shutdown sequence for that section. Sophie, start the shutdown — close the block valves at KP-127 and KP-145 remotely.
（抓起对讲机）Jorge，控制室。确认北部原油管道有泄漏，大约 KP-138处。立即派巡检队。我启动该段紧急停输程序。Sophie，开始 停输——远程关闭KP-127和KP-145截断阀。
Block valve KP-127 closing... closed. Block valve KP-145 closing... closed. Section isolated. Leak contained. Total time from alarm to isolation: 4 minutes 12 seconds. That's within our 5-minute target. I've logged the entire event in the SCADA historian — pressure, flow, and valve positions every 100 milliseconds during the event.
截断阀KP-127关闭中……已关闭。截断阀KP-145关闭中……已关闭。管段 隔离。泄漏封住了。从报警到隔离总时间：4分12秒。在5分钟目标内。 全部事件已记录到SCADA历史库——事件期间压力、流量、阀门位置每 100毫秒记录一次。