A utility-scale solar plant generates between 1 and 500 MW of power at a remote site, often 20–200 km from the nearest city. Without a working supervisory control and data acquisition (SCADA) system, the operators have no visibility into what is happening at the inverters, no way to detect a failed string combiner or a tripped feeder, and no ability to report generation data to the grid operator. SCADA is not optional at this scale. It is the nervous system of the plant, and its architecture — how many layers it has, which protocols it uses, and what data it stores — directly determines whether your operations team can resolve a fault in 15 minutes or discover it three days later when the DISCOM calls about anomalous generation figures.

Direct answer. Solar SCADA architecture for utility-scale plants follows a six-layer structure — the Solar SCADA Layer Cake: field sensors feed raw data to remote terminal units (RTUs) or data loggers, which transmit it over Modbus RTU/TCP, IEC 61850, or DNP3 protocols to a central SCADA server, displayed on an HMI console, and optionally forwarded to a cloud analytics platform. For Indian utility projects, the SCADA system must also report 15-minute generation data to the SLDC and maintain a minimum 90-day data archive. Specifying the wrong protocol or omitting cybersecurity controls are the two most common design failures.

This article is written for Suresh — an Indian utility-scale developer procuring a SCADA system for a 50 MW SECI project — and for Jennifer, a US C&I developer whose 20 MW project requires SCADA data for lender reporting. Both audiences will find the layer-by-layer specification guidance relevant, though the DISCOM reporting section applies specifically to the Indian context.

Why SCADA Is Not Optional at Utility Scale

A 50 MW solar plant has approximately 1,200 string combiner boxes, 40–100 central inverters (or several hundred string inverters), 3–5 HV power transformers, and a HV switchyard with 15–30 protection relays. Every one of these devices can fail, and every failure has a financial consequence. At ₹4.5 per kWh (a typical SECI tariff in 2025–2026), a single 100 kW string inverter that trips undetected at 08:00 and is not reset until 16:00 costs the developer approximately ₹3,600 in lost revenue on that day — and will do so every sunny day until someone physically inspects the plant.

15 min

SCADA data reporting to SLDC

CEA Grid Code 2023, India

90 days

Minimum SCADA data retention

SECI project specification, 2024

0.5%

Typical SCADA system downtime limit

IRENA best practice, 2024

1–3%

Energy loss from undetected faults

NREL utility-scale O&M study, 2023

According to NREL’s 2023 Utility-Scale Solar Operations and Maintenance report, plants with real-time SCADA monitoring detect inverter faults within 15 minutes on average, compared to 4–24 hours for plants relying on daily manual checks. The revenue impact of this detection speed difference averages 1.2% of annual generation — which at a 50 MW plant generating 90 million kWh/year represents approximately ₹60–75 lakhs in revenue at risk each year without SCADA.

Beyond revenue protection, SCADA is required by DISCOM grid codes in India for any plant connected at 33 kV and above. The CEA Grid Code 2023 mandates real-time data transmission to the State Load Despatch Centre (SLDC) at 15-minute intervals, covering active power, reactive power, voltage at the PCC, and frequency. Failure to maintain this data link can trigger PPA breach notices.

The Solar SCADA Layer Cake — Heaven Designs Architecture Framework

The Solar SCADA Layer Cake is Heaven Designs’ structured specification framework for utility-scale solar SCADA systems. It defines six layers in a bottom-up sequence, from the physical sensors at the field level to the analytics platform at the cloud level. Every layer must be specified, procured, and commissioned before the SCADA system delivers its full value.

1

Field Sensor Layer

Irradiance sensors (GHI, POA), module temperature sensors, ambient temperature sensors, wind speed and direction sensors, energy meters at combiner boxes, and protection relay IEDs. These devices generate the raw analogue and digital signals that all higher layers depend on.

2

RTU / Data Logger Layer

Remote terminal units (RTUs) or data loggers aggregate data from the field sensor layer and buffer it locally until it can be transmitted. Each RTU typically serves a cluster of inverters and one weather station. RTU selection determines which protocols are supported upstream and whether the unit can store data locally during communication outages.

3

Communication Layer

The communication infrastructure carries data from RTUs to the SCADA server. Options include fibre-optic (preferred for large sites), 4G LTE, MPLS WAN, or power-line carrier (PLC). Protocol selection — Modbus RTU over RS-485, Modbus TCP over Ethernet, IEC 61850 GOOSE, or DNP3 — must match both the RTU capabilities and the SCADA server's driver library.

4

SCADA Server Layer

The SCADA server polls all RTUs at configured intervals, stores time-stamped data in a historian database, executes control logic (such as setpoint commands to inverters), generates alarms, and feeds the HMI display. Server sizing must account for the total number of data points (tags), the polling frequency, and the required data retention period.

5

HMI Layer

The human-machine interface is the display layer that operators interact with. A well-designed HMI shows a site mimic (one-line diagram of the plant), live power output, alarm list with priority colour coding, trend charts for irradiance and performance ratio, and an event log. Poor HMI design causes operators to ignore alerts or miss critical events.

6

Analytics Cloud Layer

Cloud-based analytics platforms ingest historian data and apply performance models to calculate performance ratio, specific yield, PR degradation, and irradiance-normalised energy comparison. This layer supports lender reporting, O&M contract verification, and long-term yield bankability audits.

Field Sensor Layer — What to Specify and Where to Place It

The field sensor layer is the most under-specified part of most SCADA procurement packages. Developers often accept the inverter manufacturer’s built-in sensors as sufficient. They are not. Built-in inverter sensors measure only what is happening inside the inverter enclosure — they cannot independently verify the irradiance, the module temperature, or the wind speed that affect the performance ratio calculation.

Irradiance sensors: Specify a class A silicon pyranometer or a thermopile pyranometer for global horizontal irradiance (GHI) measurement at the weather station. For the plane-of-array (POA) irradiance measurement — which the performance ratio calculation requires — add a tilted reference cell or a second pyranometer at the module tilt angle and orientation. According to IRENA’s 2024 Solar PV Guidance for India, the POA irradiance sensor must be within 2% calibration accuracy for bankable performance guarantees.

Module temperature sensors: Pt-100 or Pt-1000 resistance temperature detectors (RTDs) bonded to the rear of a module at three representative locations in the plant. Do not rely on ambient temperature plus a nominal NOCT correction — actual module temperature depends on wind speed, soiling, and installation configuration in ways the NOCT model does not capture accurately.

Weather station: A complete weather station includes GHI pyranometer, POA pyranometer, ambient temperature sensor, relative humidity sensor, wind speed anemometer, and wind direction vane. Specify a World Meteorological Organization (WMO) class station if the plant is part of a solar resource assessment program. Mount the weather station on a dedicated mast at least 1.5 m above the module plane to avoid shading.

Energy meters at combiner boxes: Fit a DIN-rail-mounted energy meter (Class 0.5S or Class 1, per IEC 62053) at each DC combiner box. These meters capture the energy contribution from each string cluster, which makes it possible to identify a failed string or a shading anomaly at the sub-array level rather than only at the inverter level. This sub-array visibility is essential for O&M performance contracts that specify guaranteed performance ratio by cluster zone.

Definition. A data tag in SCADA is a unique identifier for one measured or calculated variable — for example, "INV-03-AC-KW" is the tag for the AC kilowatt output of inverter 3. A 50 MW plant typically has 5,000–15,000 active tags. The tag count drives SCADA server licensing costs, historian storage requirements, and communication bandwidth calculations.

RTU and Data Logger Selection

The RTU or data logger aggregates data from field sensors and inverters and forwards it to the SCADA server. RTU selection is a critical decision because the wrong choice creates protocol mismatches, communication gaps, and unsupported firmware upgrade paths that persist for the plant’s 25-year life.

Key selection criteria:

CriterionWhat to SpecifyWhy It Matters
Protocol supportModbus RTU, Modbus TCP, IEC 61850 MMS, DNP3Must match both field devices and SCADA server
Input channels8–32 analogue, 16–64 digitalDetermines how many sensors per RTU
Local data bufferingMinimum 7 days at 1-minute resolutionPrevents data gaps during communication outages
Operating temperature-20°C to +70°COutdoor cabinet temperatures in Rajasthan can reach 65°C
CertificationsCE, UL 508A, ATEX if zone-classifiedRequired for lender acceptance
Cyber hardeningNo default passwords, encrypted firmware, no unnecessary open portsIEC 62443 baseline requirement

Popular RTU and data logger platforms in the Indian utility-scale market include Satec, Janitza, Schneider Electric EcoStruxure, Socomec Diris, and ABB’s RTU560 series. The inverter manufacturer’s proprietary data logger (such as SMA’s Sunny Portal gateway or Huawei’s SmartLogger) can serve as a supplementary device but should not be the sole data collection point because manufacturer loggers may not report to third-party SCADA systems over open protocols.

Field tip. Always request a Modbus register map from the inverter manufacturer before finalising the RTU specification. Some manufacturers provide read-only Modbus access on their standard product; others require a paid gateway licence to enable third-party SCADA communication. Discover this cost at the specification stage, not after the inverters are delivered on site.

Communication Protocols — Modbus, IEC 61850, and DNP3

The communication protocol is the language that RTUs and SCADA servers use to exchange data. Choosing the wrong protocol — or mixing incompatible protocol versions — is the single most common SCADA commissioning failure in Indian utility-scale projects. The three dominant protocols in solar SCADA are Modbus, IEC 61850, and DNP3.

Modbus RTU and Modbus TCP: Modbus is the most widely supported protocol in solar inverters, energy meters, and RTUs. Modbus RTU operates over RS-485 serial links; Modbus TCP operates over standard Ethernet. Modbus is simple and reliable but has no built-in authentication, no data encryption, and no time-synchronisation mechanism. It is appropriate for the field-to-RTU layer but not for internet-facing connections.

IEC 61850: IEC 61850 is the standard for communication in electrical substations, defined by the International Electrotechnical Commission. It provides a structured data model (logical nodes), time-stamped messaging (GOOSE for fast events, MMS for sampled values), and a configuration language (SCL files) that enables interoperability between IEDs from different manufacturers. According to IEEE C37.1 (Standard for SCADA and Automation Systems), IEC 61850 is the preferred protocol for HV protection relay communication in solar plants above 10 MW.

DNP3: DNP3 (Distributed Network Protocol version 3) is widely used in USA utility SCADA systems for communication between remote outstations and the control centre. It includes time-synchronisation, data integrity checking, and unsolicited reporting — features that Modbus lacks. For Indian projects connecting to SLDC SCADA systems, DNP3 may be required if the SLDC specifies it in the grid connectivity documentation.

ProtocolLayerSpeedSecurityIndia UseUSA Use
Modbus RTUField to RTU9.6–115.2 kbpsNoneInverter to RTUInverter to RTU
Modbus TCPRTU to SCADA10–1,000 MbpsNoneLAN SCADALAN SCADA
IEC 61850 MMSSubstation to SCADA10–1,000 MbpsTLS optionalHV protection IEDsHV protection IEDs
IEC 61850 GOOSEIED to IEDunder 4 msNoneTrip commandsTrip commands
DNP3RTU to control centre9.6 kbps–100 MbpsSA5 extensionSLDC reportingUtility SCADA

SCADA Server Architecture and Sizing

The SCADA server is the central data processing engine. Sizing it correctly requires knowing the tag count, the polling interval, the data retention period, and the redundancy requirements.

Redundant server topology: For a utility-scale project under a SECI PPA, the SCADA server must have hot-standby redundancy — two servers that mirror each other in real time, so that a server failure does not interrupt data recording or SLDC reporting. The switchover time from primary to standby must be under 30 seconds to avoid a reportable data gap. This requirement doubles the hardware cost but is non-negotiable for projects with lender oversight.

Historian database sizing: The historian stores time-stamped values for every tag at every scan cycle. A 50 MW plant with 10,000 tags polled at 1-minute intervals generates approximately 14.4 million data records per day, or 5.2 billion records per year. At 50 bytes per record (timestamp + tag ID + value + quality flag), this requires approximately 260 GB of storage per year. Specify a minimum 5-year storage capacity to avoid mid-life disk replacement during the revenue-generating period.

OPC-UA interface: The SCADA server must expose an OPC-UA (Open Platform Communications Unified Architecture) interface for integration with the analytics cloud layer, the inverter manufacturer’s monitoring platform, and any lender-appointed monitoring systems. OPC-UA is the ISO/IEC 62541 standard for industrial data exchange and is the protocol most commonly required by lenders for third-party read-only access to plant data.

Watch out. A single-server SCADA architecture at a utility-scale plant creates a single point of failure for both operations and DISCOM reporting. If the server fails during a grid disturbance, the plant has no visibility into protection relay operations at the exact moment it matters most. Specify redundant servers and a UPS rated for at least 4 hours of operation.

HMI Design Requirements

The HMI is the interface that determines whether operators act on alarms quickly or ignore them. Poorly designed HMIs contribute to alarm fatigue — a condition where operators are presented with so many simultaneous alarms that they stop responding to individual ones. According to IRENA’s Solar PV Operations and Maintenance Best Practices (2022), plants that redesigned their HMI alarm hierarchy after commissioning reduced alarm response time by 40% on average.

The HMI design must follow these principles:

  • Site mimic as the home screen: The home screen must show a one-line diagram of the plant, colour-coded by health status (green = normal, amber = warning, red = fault). Operators must be able to drill down from the site-level view to a cluster view, then to an inverter view, then to a string combiner view with no more than three clicks.
  • Alarm prioritisation: Alarms must be classified into four levels: Level 1 (critical — plant tripped, SCADA server failure, SLDC communication loss), Level 2 (high — inverter trip, breaker trip, relay operation), Level 3 (medium — sensor failure, communication timeout), Level 4 (low — data quality flag, minor deviation).
  • Performance ratio dashboard: A live performance ratio (PR) calculation — comparing actual energy output to the irradiance-normalised expected output — must be visible on the HMI home screen. A PR below 70% for more than 30 minutes during daylight hours should trigger an automatic Level 2 alarm.
  • Trend charts with zoom: Operators must be able to view trend charts for any tag over any time period from 1 hour to 12 months. This capability is essential for diagnosing intermittent faults — a string combiner that trips for 10 minutes every afternoon due to a loose connection will only be visible on a multi-day trend chart.

DISCOM and Grid Operator Data Reporting in India

For Indian utility-scale projects, the SCADA system is not solely an internal operational tool — it is also a regulatory reporting instrument. The CEA Grid Code 2023 and the state-specific grid codes require the plant SCADA system to transmit real-time data to the SLDC at 15-minute intervals. The data points required are:

  1. Active power at the PCC (MW)
  2. Reactive power at the PCC (MVAr)
  3. Voltage at the PCC (kV)
  4. Frequency (Hz)
  5. Plant availability status (available, unavailable, or partially available)
  6. Reason for unavailability if applicable

The communication link to the SLDC is typically established over a dedicated leased line (MPLS or ISDN) provided by the state transmission company. The data format is usually IEC 60870-5-101 (serial) or IEC 60870-5-104 (TCP/IP). The developer must provide the SCADA interface specification to the SLDC’s dispatch engineering team and obtain written confirmation of the interface protocol before commissioning begins.

Definition. SLDC (State Load Despatch Centre) is the body responsible for scheduling, dispatch, and grid security management within a state. For solar projects, the SLDC uses the 15-minute generation data from the plant SCADA to verify that the plant is following its generation schedule and to settle imbalance charges under the deviation settlement mechanism (DSM).

See our detailed guide on the DISCOM net metering process across Indian states for state-specific SCADA and data reporting requirements that vary from the CEA national baseline.

Cybersecurity Considerations for Solar SCADA

Solar plant SCADA systems were historically isolated from the internet and therefore not considered cybersecurity risks. That assumption is no longer valid. Most modern SCADA systems have internet-facing connections for remote monitoring, cloud analytics, and remote inverter firmware updates. A cyberattack on a utility-scale solar plant can cause uncontrolled disconnection from the grid.

According to CEA’s Cyber Security Guidelines for Power Sector 2021, utility-scale solar plants must implement the following baseline cybersecurity controls:

  • Network segmentation: The OT network (SCADA, RTUs, protection IEDs) must be physically separated from the IT network using a data diode or industrial firewall. No bidirectional connection between OT and IT networks is acceptable.
  • Access control: All SCADA system access must require individual username and password authentication. Default passwords on any device must be changed before commissioning. Privileged access (control commands) must require two-factor authentication.
  • Patch management: SCADA software and firmware must be updated within 30 days of security patches being released by the vendor. A formal patch management procedure must be documented and audited annually.
  • Incident response plan: The plant operator must maintain a documented cyber incident response plan that defines who is notified, in what order, within what timeframe, when a security breach is detected.

OPEN PROTOCOL ADVANTAGES

  • Interoperability with any SCADA platform
  • Multiple vendor options for RTU and server
  • Third-party lender monitoring access without proprietary licence
  • Long-term vendor independence over the 25-year plant life

PROPRIETARY SYSTEM RISKS

  • Locked to single inverter manufacturer's ecosystem
  • No third-party read access without manufacturer permission
  • Vendor discontinues support mid-plant-life
  • Cannot integrate equipment from other brands during plant expansion

Download a sample solar SCADA specification

Includes tag list template, RTU specification checklist, HMI design brief, and SLDC reporting interface specification for a 50 MW Indian utility project.

Get the sample pack →

How Heaven Designs Helps with Solar SCADA Engineering

SCADA specification is not a task that most EPC contractors handle well. It requires electrical engineering knowledge (protocol selection, CT/VT signal conditioning), software engineering knowledge (database sizing, historian configuration), and regulatory knowledge (SLDC reporting requirements, CEA cyber guidelines). Assembling this expertise in-house for a single project is expensive and creates a capability that sits idle between projects.

Heaven Designs supports utility-scale developers and EPCs with SCADA engineering from the specification stage through commissioning acceptance testing. Our work integrates with the broader solar engineering workflow for Indian EPCs so the SCADA specification is consistent with the HV drawings, relay coordination study, and IFC drawing package.

Contact us to discuss your SCADA engineering requirements. Heaven Designs has delivered SCADA specifications for projects in Rajasthan, Gujarat, Andhra Pradesh, and Karnataka accepted by both independent engineers and SLDC interface teams.

FAQ

What is the difference between a SCADA system and an inverter monitoring portal?

An inverter monitoring portal (such as SMA’s Sunny Portal or Huawei’s FusionSolar) is a proprietary cloud platform provided by the inverter manufacturer. It displays data from that manufacturer’s inverters but typically does not integrate data from protection relays, energy meters, weather stations, or other equipment brands. A full SCADA system aggregates data from all equipment in the plant across brands and device types, provides real-time alarming, historian storage, SLDC reporting, and control capabilities. For utility-scale projects under regulatory obligations, the inverter portal alone is not a SCADA system.

How many data tags does a 50 MW solar plant typically have?

A 50 MW ground-mount plant with 40 central inverters, 120 combiner boxes, one weather station, and one HV switchyard typically has between 8,000 and 15,000 active SCADA tags. This includes 50–80 tags per inverter, 8–12 tags per combiner box, 20–30 tags for the weather station, and 200–400 tags for the HV protection relays. The tag count drives SCADA server licensing, historian storage sizing, and communication bandwidth requirements.

Is Modbus TCP secure enough for solar SCADA communication?

Modbus TCP has no built-in security features — it does not authenticate clients, encrypt data, or detect data integrity violations. Within a protected LAN environment (isolated OT network behind an industrial firewall), Modbus TCP is acceptable for field device communication. It must not be used for any internet-facing communication. For those connections, use OPC-UA with TLS encryption or HTTPS-based REST APIs with certificate authentication.

What is the SLDC reporting requirement for a SECI project in India?

Under the CEA Grid Code 2023, the plant SCADA system must transmit active power, reactive power, voltage, and frequency data to the SLDC at 15-minute intervals continuously during the plant’s operating life. The communication link is typically IEC 60870-5-104 over a dedicated leased line. The SLDC uses this data for generation scheduling and deviation settlement. Data gaps exceeding 30 minutes may trigger a deviation settlement penalty under the PPA.

How do I calculate SCADA historian storage requirements?

Calculate the number of active tags multiplied by the polling frequency multiplied by the bytes per data record multiplied by the retention period. For example: 10,000 tags times 1 scan per minute times 50 bytes per record times 525,600 minutes per year equals 262.8 GB per year. For a 5-year retention requirement, specify at least 1.5 TB of usable storage with RAID-6 redundancy. Add 50% overhead for database indexing, operating system, and application software.

Can a solar SCADA system send control commands to inverters remotely?

Yes. Modern SCADA systems can send active power setpoint commands (in kW or as a percentage of rated power), reactive power setpoints (in kVAr or power factor), and on/off commands to inverters over Modbus TCP or IEC 61850. This capability is required for plants that must participate in grid frequency response or that receive curtailment orders from the SLDC. The control logic must include safeguards to prevent a software error from sending an invalid setpoint that damages equipment or causes a grid disturbance.

What cybersecurity standard applies to solar SCADA in India?

CEA’s Cyber Security Guidelines for Power Sector (2021) apply to all grid-connected generation plants. For internationally funded projects, lenders may additionally require compliance with IEC 62443 (Industrial Automation and Control Systems Cybersecurity). The minimum requirements include network segmentation between OT and IT, access control with individual user accounts, patch management procedures, and an incident response plan. SECI projects are increasingly requiring a cybersecurity audit by an independent third party before commissioning approval is granted.