The monitoring paradox
Every solar plant in India today has some form of remote monitoring. Data is being collected: inverter parameters, irradiation levels, energy generation, grid export. Alarms are being triggered. Dashboards are being looked at.
And yet, across most of the mid-to-small IPP segment, plant operators are still being surprised by underperformance. Generation shortfalls are discovered days after they begin. Equipment degradation is caught after it has already cost significant yield. Portfolio-level patterns that should be obvious go unnoticed because nobody is looking at the data in the right way.
Indian solar is, in large part, data rich and insight poor.
The problem is rarely the absence of an RMS. The gap lies between what most RMS implementations are doing and what they are capable of doing. Most solar IPPs are drawing on roughly 20-30% of the value their monitoring infrastructure could deliver. The remaining 70-80%, in the form of better decisions, earlier interventions, lower losses and more credible reporting, sits untapped in data that is already being collected.
What most RMS implementations actually do
At the basic end, which is where most implementations sit, an RMS collects data from inverters, weather stations and meters at defined intervals, displays it on a dashboard and generates alerts when a parameter crosses a threshold. Generation is tracked, availability is logged and a monthly report is produced.
This is genuinely useful. But it has several significant limitations:
- Data without context: Raw generation numbers mean little without comparison to expected generation based on irradiation. An RMS that tells you a plant generated 10,000 units yesterday is informative. An RMS that tells you the plant should have generated 11,500 units based on the day's irradiation and therefore underperformed by 13% is actionable.
- Alarms without intelligence: Most RMS implementations generate alarms when a parameter crosses a threshold but do not distinguish between alarms that require immediate intervention and alarms that are noise. Without prioritisation and pattern recognition, alarm fatigue sets in and critical signals get lost in the volume.
- Device-level data without fleet-level insight: An RMS that monitors one plant in isolation gives you one plant's data. An IPP with multiple plants needs fleet-level visibility to compare performance across plants and identify systemic patterns.
- Periodic reporting without real-time decision support: Most RMS-generated reports are produced monthly or weekly, too slow for operational decision-making. By the time a monthly report flags a performance issue, weeks of generation loss may have already occurred.
The hidden cost of the gap
Consider a 10 MW solar plant with a Performance Ratio of 75% when it should be achieving 80%. That 5 percentage point gap translates to roughly 2.5 to 3 lakh units of lost generation annually at typical Tamil Nadu irradiation levels. At a PPA tariff of Rs 3.50 per unit, that is Rs 8 to 10 lakhs of revenue lost every year. Over a 25-year plant life, the compounding effect is significant.
Beyond direct generation loss, there are less visible costs: higher O&M spend from reactive rather than proactive maintenance, warranty and insurance leakage from gaps in monitoring records and reduced investor and lender confidence from manually compiled reports.
Six areas where the gap is widest
1. Telemetry: the foundation everything else depends on
Before any discussion of analytics, KPIs or forecasting, there is a more fundamental question that most RMS implementations never ask clearly enough: is the data being collected actually reliable?
Telemetry, the infrastructure that connects physical equipment on the plant to the monitoring system, is the foundation on which everything else is built. Sensors, data loggers, communication protocols, network connectivity, polling frequencies, signal integrity: these are not glamorous topics. They rarely feature in RMS vendor presentations. But they determine, more than any other factor, whether the data flowing into your monitoring system is trustworthy.
The consequences of poor telemetry are insidious precisely because they are invisible. A plant with unreliable telemetry does not announce itself. It produces data that is quietly wrong. Readings drop out intermittently. Parameters flatline during communication failures. Timestamps drift. Each anomaly looks like noise; in aggregate, they corrupt every metric, every KPI and every insight the RMS generates downstream.
In our own experience building monitoring infrastructure from the ground up, the importance of getting telemetry right before investing in any analytics capability was one of our most significant early learnings. A sophisticated analytics layer built on unreliable telemetry produces sophisticated wrong answers.
Practically, this means paying serious attention to:
- Sensor selection and placement: using calibrated, appropriately rated sensors positioned correctly for the measurement they are making
- Data logger specification: ensuring loggers have sufficient channel capacity, appropriate polling frequencies and reliable local storage for communication outages
- Communication architecture: designing for redundancy, not just connectivity, so that a single point of failure does not create a data black hole
- Protocol standardisation: choosing communication protocols appropriate for the equipment being monitored and the data volumes being handled
- End-to-end testing: validating the full data chain from sensor to dashboard before commissioning, not after
Telemetry is rarely value-engineered out of project budgets deliberately; it tends to be the item that slips quietly. Retrofitting it after commissioning is significantly more expensive and disruptive. Getting it right at design stage is an operational discipline decision that pays for itself many times over.
2. Data quality and communication health
Even with well-designed telemetry, communication failures create gaps in the data record that distort performance calculations and hide real problems. Most basic RMS implementations log communication failures but do not actively manage data quality: filling gaps intelligently, flagging suspicious readings or distinguishing between a genuine underperformance event and a data collection failure. Building this layer in is foundational to everything that follows.
3. KPI computation and benchmarking
Performance Ratio, Capacity Utilisation Factor, Plant Load Factor, Specific Yield, Availability: these are the metrics that define how well a plant is performing. Most RMS systems display raw generation data but do not automatically compute these KPIs, compare them against design expectations or benchmark them against similar plants. Automating them, with clear visibility into what is driving deviations from expected values, is one of the highest-value improvements an IPP can make to its monitoring stack.
4. Yield forecasting
Knowing how much a plant generated yesterday is history. Knowing how much it is likely to generate tomorrow, based on weather forecasts, historical irradiation patterns and current equipment health, is planning. Yield forecasting enables better SLDC scheduling, more accurate energy banking management and proactive communication with offtakers. Most mid-to-small IPPs have no systematic yield forecasting capability: generation is tracked retrospectively, not anticipated.
5. Multi-OEM data normalisation
Most IPP portfolios include equipment from multiple manufacturers, each using different data formats, parameter names and communication protocols. Without a normalisation layer that translates all of this into a common data model, fleet-level analytics are impossible. Every plant becomes a data silo. An IPP that can view all its plants through a single normalised data lens has a fundamentally different analytical capability from one managing each plant's data separately.
6. Revenue meter integration
The revenue meter, the interface between the plant and the grid, is the most financially consequential piece of equipment on any solar plant. Discrepancies between inverter-reported generation and revenue meter readings are common, often unexplained and frequently unresolved. Integrating revenue meter data directly into the RMS with automated reconciliation against inverter data closes this loop and provides an auditable, defensible record of actual generation for PPA billing and regulatory compliance.
What it takes to go further
Closing the gap between basic monitoring and genuine plant intelligence is an architectural challenge more than a technology one. The tools exist. The data is already being collected. What is needed is a monitoring system designed from the ground up to deliver insight, not just data.
The IPPs that get the most value from their monitoring infrastructure treat their RMS as a strategic asset. They invest in telemetry and data quality before analytics. They think at fleet level. They build for the full asset lifecycle. At Good Energies, this has been a priority from early in our journey. We are building an in-house RMS with telemetry architecture, KPI computation, data quality management, yield forecasting, multi-OEM normalisation and revenue meter integration as foundational capabilities, with further modules covering the full plant lifecycle in development. We are not where we want to be yet, but we know exactly where we are going and why it matters.
The opportunity ahead
The Indian solar industry has done a remarkable job of scaling generation capacity over the past decade. The next decade will be defined not by how much capacity gets added, but by how well that capacity is operated. The gap between a solar plant that is monitored and a solar plant that is intelligently managed is a mindset gap: a decision to treat data as a strategic asset rather than a reporting obligation. For C&I buyers evaluating long-term PPA partners, the quality of an IPP's monitoring and asset management infrastructure is an increasingly important signal of operational credibility and of whether the power supply they are committing to for the next decade will actually be there when they need it.