Power and renewable projects are producing far more data than they used to, and the control systems that collect it have to scale without letting that data run wild. In a session titled “Big Data, Big Solutions: Keep it Under Control,” Matt Roberts, Director, NA Renewables Business Development at Emerson, walked through how the Ovation platform has grown to meet that demand and how engineers can plan systems that stay reliable as point counts climb.

Why It Matters

A 100 MW gas turbine plant might carry a few hundred I/O points, and use roughly 200 controller points per megawatt, with operation handled mostly on site. Renewable and hybrid sites are a different story: solar, wind, and battery projects run up to 2,000 points per megawatt, communicate mostly through datalinks, and operate remotely from a central location.

Matt described solar sites now discussed at 5 gigawatts, battery projects at 30 gigawatt-hours, and behind-the-meter data center projects reaching 10 gigawatts, with fleets of hundreds of sites and tens of millions of points. Every indicator points to that trend continuing, which means planning matters more than raw capacity.

Key Takeaways

• Data volume is climbing fast. Ovation 4.0 now handles 750,000 points per second in the hardware space with a gigabit network, up from 200,000 in the 3.6 generation, and Ovation Green SCADA has been validated to more than 4 million points.
• Most points are not control points. In a sample 50 MW solar plus 25 MW battery site, controls made up under 5 percent of the data, with condition monitoring at 35 percent and statuses and faults at 55 percent.
• Four elements drive scaling. Device count, point count, scan rate, and configuration time each create their own technical and practical constraints.
• Configuration time is the quiet risk. Managing changes across thousands of near-identical devices is often the hardest part of repeatable project execution.
• The right answer is architecture-specific. Stated limits reflect validated capacities, not guarantees, so every large system needs engineering review for its specific requirements.

Sort the Data Before You Scale the System

The core message was to stop treating all data the same. When a site has only a few thousand points, everything can go into one bucket at one rate. At scale, that breaks down. Matt categorized every point on a real hybrid site and found that out of 24,000 points, only about 600 needed to sit in closed-loop logic. The rest were monitoring, condition, or fault data that can be routed and stored differently.

That distinction lets engineers match each stream to the right path and the right scan rate, from a solar tracker that updates slowly to a battery inverter or turbine controller that needs frequent reads. Getting this planning right is the difference between a bigger bucket and the right buckets.

Know the Building Blocks and Their Limits

The platform gives several ways to get data in, each with defined capacities. A single Ovation Highway, supported by its database server, carries up to 750,000 points per second, with up to 254 drops. Controllers range from 10,000 points on the OCC100 to 64,000 on the OCR3000 and SDC100. For high-volume datalink traffic, an Ovation SCADA Communication Server handles up to 128,000 points, can run in a redundant configuration, and can even feed the historian directly, bypassing the highway when point counts get tight.

For the fastest data, an ELC (Ethernet Link Controller) module supports up to 100 Modbus devices, with up to 16 modules and 20,000 points per controller. Newer embedded CPS (Communication Protocol Suite) connects up to 256 devices per controller directly from the hardware, using the same method as software-defined control. On the storage side, the OPH (Ovation Process Historian) holds up to 500,000 points, at 50,000 points per scanner.

Extend Across Sites Without Overwhelming the Network

For fleets, Ovation Multi-Network ties systems together through a data access server at each site, connecting up to 300 networks and a theoretical 225 million points. The design keeps data at the edge and pulls it centrally only when needed.

As Matt and some of the attending Ovation subject-matter experts explained, an operator workstation subscribes to remote points only while a graphic, trend, or point display is open, then unsubscribes when it closes. That approach keeps traffic manageable over lower-bandwidth radio or cellular links, where paying per gigabyte makes streaming every point every second impractical. Multi-network also serves as a soft boundary between systems, useful for reliability isolation or keeping a smaller site from inheriting the regulatory obligations of a larger block.

Optimize Points and Ease Replication

Once the architecture is set, disciplined configuration keeps it lean. Scan types can run periodically, by exception, on trigger, or by mapping, and Modbus read scanblocks can span up to 125 registers to maximize throughput per call. OPH-only points route straight to the historian without counting against network limits, which helps trends and reports that never touch logic.

Practices like packing default digital points, using the SCADA 226 command to set datalink values without extra logic, and the VCALC command for complex math all trim overhead. For repeatable work such as solar and battery builds, control macros and Object Builder let teams define a block once, instance it many times, and push updates from an abstracted layer rather than sheet by sheet. These methods are especially valuable for the highly repeatable designs common in solar power plant controls.

Plan for What Comes Next

The direction is clear: validated limits keep expanding, embedded CPS and software-defined control keep maturing, and configuration tools keep cutting the time it takes to build and maintain these systems. The engineers who benefit most are the ones who plan ahead, use their space efficiently, and clean up after themselves.

To see how the platform supports projects from the field to the cloud, explore the Ovation Automation Platform.