In the digital build-out of a new-generation renewable energy centralized control center, an integrated wind-solar-storage plant is In the digital build-out of a renewable energy control center, an integrated wind-solar-storage plant is no longer just a scattered set of grid-connected equipment nodes. It is a tightly coupled system in which solar irradiance, equipment status, electrochemical safety, step-up power supply, and grid dispatch metrics are deeply interwoven.
For the operations staff at the centralized control center, data is never scarce. The real challenge is that when hundreds of wind turbines, inverters, converters, and battery management systems produce massive amounts of data in real time, the field still struggles to answer a few key business questions quickly. Where does the faint precursor of an equipment anomaly lurk? What chain reaction will it cause to generation output and battery stack safety? Are we facing a single-unit electrical disturbance, or a major system risk that already threatens the grid assessment indicator (AGC deviation) and the overall availability of the plant?
This is the critical dividing line for renewable energy control center operations. Operators need more than equipment parameters stacked on a big screen. They need data that can reconstruct the causal propagation chain running from wind-solar generation → battery storage → main transformer step-up → grid interconnection gateway, guiding the site from passive monitoring to precise early warning and coordinated decision-making.
Renewable energy control: operating a tightly coupled wind-solar-storage system
The assets under a renewable energy control center are widely distributed and varied, covering core systems such as wind farms, PV plants, energy storage stations, step-up substations, and grid interconnection gateways. Tight physical and dispatch coupling exists between generation equipment (such as wind turbine WT-T01 and inverter PV-INV01), storage systems (converter ES-PCS01 and battery management system ES-BMS01), transformers, and energy meters.
Under normal combined wind-solar-storage operation, the balance between systems is highly dynamic. The instantaneous output of wind or solar determines how well the transmission link is fed. If an output gap appears, the centralized control center must dispatch the storage converter to discharge quickly to fill the shortfall. Meanwhile, sustained high-rate discharge of the storage battery directly accelerates cell temperature rise and aging, and can even trigger thermal runaway.
In this physical network of multi-energy complementarity and interlocked systems, traditional single-point equipment monitoring cannot see the whole picture. When the grid-port meter detects that AGC deviation exceeds the limit, operators often have to move across the big screens of several subsystems to investigate, and it is hard to quickly confirm whether wind turbine bearing wear triggered automatic power limiting, or whether the battery stack ran short of charge and dropped out of dispatch compensation.
Too many alarms, slow diagnosis: the pain points of operating a tightly coupled plant
The same fluctuation can be monitored, but the cause is still hard to pin down
At the central control site, teams often hit the bind where the fluctuation is visible but the root cause is unclear. For example, a drop in a wind turbine’s generation output could be weaker wind, or it could reflect deteriorating mechanical efficiency and SCADA derating protection caused by poor gearbox lubrication. If they only watch the power reading, operators cannot make a precise call in the first instance; they must cross-compare the reading against contextual data such as lubricating oil pressure, bearing vibration, and generator winding temperature.
A single-point alarm has triggered, but the chain effect is hard to judge
Fault propagation in renewable energy equipment shows the classic cross-device cascade pattern. Once the insulation resistance on the DC side of a PV inverter drops from moisture, leakage current rises, MPPT tracking shifts, inverter efficiency suffers, and derating protection triggers. The resulting charge shortfall further pushes the storage battery stack to hit the 10% discharge cutoff early during nighttime discharge, triggering a protective shutdown and finally a sharp drop in the step-up transformer’s load factor. In a traditional SCADA system, these alarms are isolated across different subsystems with no chain linking them, so operators delay their decisions when faced with a flood of alarms.
The business impact is not abstract: assessment fines and battery stack damage are costly
The cost of a delayed response is high. Two scenarios illustrate the point:
- In the wind turbine gearbox wear case: a seized lubrication pump on the turbine increases mechanical friction, and the SCADA system derates the turbine’s active power to 800 kW (normal is 1650 kW) to protect the equipment. Grid-port AGC deviation worsens from a normal 1.2% to -5.5%, beyond the ±3% assessment limit allowed by the grid, exposing the plant to heavy grid assessment fines. At the same time, the storage battery is forced to discharge at high power to cover the gap, cell temperature rises to 38°C and a thermal runaway warning triggers, shortening battery life.
- In the inverter insulation dampness case: inverter power limiting means the charging plan is not completed, so storage battery stack No. 2 discharges to the 10% cutoff voltage overnight and drops out of service. The main transformer load factor falls from 45% to 22%, and the generation plan misses badly.
From passive monitoring to proactive judgment: reorganizing the data chain
Object-based modeling: scattered measurement points return to a unified business view
To answer where an anomaly happens and how it affects the whole system, how the data is organized is key. TDengine uses Industrial Ontology modeling to map more than ten thousand sensor measurement points across wind, solar, storage, step-up substations, and grid interconnection gateways into a clear tree-shaped asset system.
Figure 1: A tree hierarchy organizes plant assets and measurement points into a unified business view
Under this model, every piece of data is bound to a specific equipment entity and process node. Operators no longer face a chaotic signal stream but equipment objects with physical hierarchy and topology, which provides the logical backbone for later fault propagation tracing.
Real-time analysis and event linkage: alarms return to full process context
On top of the asset model, TDengine’s real-time analysis engine continuously runs real-time aggregation on the wind-solar-storage data stream. When it detects a measurement point crossing its limit or an operating anomaly, it dynamically generates an event context and pushes the anomaly duration, rate of change, and affected neighboring parameters together to the duty desk.
TDengine also provides Chat BI and proactive insight capabilities. When the grid-port meter triggers an AGC deviation alarm, central control staff do not need to write complex SQL. A simple plain-language query lets the system compare the current wind turbine output and storage converter action and recommend the best primary and standby adjustment plan.
Figure 2: General information settings for real-time analysis
Figure 3: Trigger conditions for real-time analysis
Process analysis and AI-assisted insight: judgment shifts from experience-driven to evidence-driven
Once multi-source time-series data is fused and linked by object, fault investigation at the centralized control center shifts from experienced staff making educated guesses to data-driven investigation aligned on a time axis.
The system provides intelligent process analysis and multi-dimensional shift comparison. When facing complex situations such as inverter power decay or abnormal battery SOC, central control staff can start an AI-assisted interpretation with one click. The AI-assisted analysis overlays the current operating curves on historical golden reference segments, quantifies the deviation, traces back the precursor features, and produces a fault report with root-cause conclusions and time-sequenced evidence, guiding maintenance crews to go out precisely.
Figure 4: AI interpretation and data mining on the analysis panel
Figure 5: AI root-cause analysis of an event
The analysis loop in typical anomaly scenarios
Anomaly case 1: progressive wind turbine bearing wear leads to an output gap and emergency storage compensation
In dispatch period H-048 to H-067, the gearbox lubricating oil pressure of wind turbine WT-T01 first fluctuated down from a normal 3.2 bar to 1.6 bar (below the 2.1 bar Warning line, lasting 5 minutes). Because of poor lubrication, the gearbox bearing wear accelerated over the following hours. Bearing vibration jumped from 1.2 mm/s to 8.5 mm/s (above the 4.0 mm/s Major alarm line), gearbox oil temperature rose to 75°C (above the 72°C Major alarm line), and generator winding temperature climbed to 118°C. SCADA triggered derating protection and active power was forced down from 1650 kW to 800 kW. Converter No. 1 at the storage station detected the output gap and discharged 3200 kW to compensate, driving battery stack No. 1’s SOC from 55% down to 18% and firing a thermal runaway warning. In the end, grid-port AGC deviation still worsened to -5.5%, causing the dispatch assessment to be exceeded.
After the grid-port meter detected the AGC deviation exceedance, the operations team at the centralized control center quickly ran the following three-step trace:
Step 1: Confirm the derating response of the generation equipment.
Check the active power and operating status curves of WT-T01. The power falling sharply from 1650 kW to 800 kW shows the turbine itself triggered protective derating and curtailment because its own mechanical temperature or vibration exceeded limits, not because outside wind weakened.
Step 2: Assess the degree of mechanical and thermal deterioration.
Overlay the gearbox vibration, gearbox oil temperature, and generator winding temperature curves of WT-T01. The data shows vibration swinging violently in the high range around 8.5 mm/s, gearbox oil temperature holding at 75°C, and generator temperature reaching 118°C. This indicates serious metal friction and heat buildup inside the gearbox; the mechanical system is already in a deteriorating state.
Step 3: Trace the underlying root cause and quantify the system loss.
Drill down into the lubricating oil pressure parameter, which confirms it fell below 1.6 bar tens of minutes before the vibration climbed sharply, exposing the root trigger of the seized lubrication pump motor. At the same time, trace the SOC drop slope and thermal runaway indicators of battery stack No. 1 to quantify the cell-life loss and economic cost (the grid-port AGC assessment exceedance) that the storage system paid to fill this gap.
Figure 6: Correlation analysis of wind turbine fault and energy storage emergency compensation
Through this chain, the root cause is confirmed: wear and seizing of the lubrication pump motor → a sharp drop in gearbox lubricating oil pressure → intensified bearing friction and vibration beyond limits → rising generator temperature and derating protection → emergency high-rate storage discharge to compensate → rising battery cell temperature and grid AGC deviation exceeding the limit.
Anomaly case 2: damp DC-side inverter insulation causes leakage and a protective battery shutdown
In dispatch period H-088 to H-107, the rubber seals of the DC combiner box aged, inverter PV-INV01 became damp, insulation resistance dropped sharply from 18500 kΩ to 450 kΩ (below the 2000 kΩ Major alarm), and leakage current rose from 4.2 mA to 52 mA (above the 25 mA Warning alarm). The degraded insulation let DC-side energy dissipate: inverter efficiency fell from 98.0% to 93.8% and output power from 810 kW to 520 kW (below the 600 kW Major alarm). Because of the PV output gap, battery stack No. 1 charged to only 62% during the day (planned 78%), so during nighttime discharge battery stack No. 2 was drained to 10% early in the morning and triggered a protective shutdown. The load factor of main transformer No. 1 then fell from 45% to 22%.
Central control operations staff traced the fault chain through the process workbench around the storage shutdown and transformer unloading event:
Step 1: Monitor the electrical insulation and leakage indicators.
Watch the insulation resistance and leakage current trends of PV-INV01. Resistance slid down to 450 kΩ in a short time while leakage current climbed to 52 mA. This confirms a serious leakage path to ground on the combiner box DC side, where electrical energy never enters the inverter circuit and is instead turned into heat loss in the resistance to ground.
Step 2: Assess the inverter conversion efficiency and output deterioration.
Compare the inverter DC-side voltage, conversion efficiency, and active power output. DC voltage is pulled down to 662 V by the insulation leakage, efficiency drops to 93.8%, and active power output falls to 520 kW, directly forming a generation output gap.
Step 3: Quantify the storage shortfall and the final impact on the main transformer side.
Trace the charge and discharge SOC curves of each battery stack at the storage station. Stack No. 1, short on daytime charging, cannot deliver enough discharge power at night, so stack No. 2 is drained to 10% early and drops out of service. The system performs a protective shutdown for safety, which directly drops the main transformer load factor to 22%, and the main grid side faces penalty risk from insufficient output.
Figure 7: Leakage current event propagation across inverter, battery, transformer, and insulation variables
This case fully reconstructs the following evolution: poor combiner box sealing lets moisture in → DC-side insulation resistance to ground falls and leakage current rises → DC bus voltage is pulled down and inverter efficiency drops → the inverter derates its power output → a planned gap appears in daytime storage charging → storage hits the limit early during nighttime discharge and shuts down protectively → the step-up station main transformer load factor drops sharply.
Renewable energy control needs connected judgment across energy systems
As integrated wind-solar-storage plants expand, the core value of an intelligent system is no longer providing big screens full of sensor readings. It is turning these time-series streams into decision support for plant operations and dispatch safety.
TDengine maps multi-dimensional source data into clear equipment object models, binding wind-solar generating units, battery stack safety, and grid dispatch together, and gives renewable energy control centers a trusted diagnosis chain that helps plants supply the grid steadily and reliably.
TDengine comes with a high-performance, distributed time-series database, Industrial Ontology modeling and an Industrial Agent Runtime, providing a full-stack solution for industrial data streams from collection and storage to real-time analytics, visualization, event management, and root-cause analysis. To learn more about TDengine, visit www.tdengine.com and download it free.
Try it yourself
Install and deploy TDengine Visit the TDengine Download Center, select TDengine All-in-One, choose the deployment platform and architecture that matches your environment, and follow the guided steps to complete the installation.
Load the sample data
On first activation, on the sample data loading screen, select New Energy Centralized Control Center and wait for loading to complete.
If you have already activated the product, click your avatar in the top-right corner, select the Management Console, choose Sample Data on the left, then select New Energy Centralized Control Center to load it. Wait a few minutes for loading to finish, and you are ready to explore.
New Energy Centralized Control Center
