Your Solar and Storage Monitoring Portals Don’t Tell the Whole Story. Data Assembly Completes the Picture

Your Solar and Storage Monitoring Portals Don't Tell the Whole Story

After installing solar and storage infrastructure, providers hand over a portal for monitoring the system. But that portal only furnishes a slice of the information needed to accurately assess the performance of these assets. The rest of the data sits across about seven separate locations.

Only by assembling all these pieces into a centralized platform can organizations make sense of whether their equipment is operating optimally, billing and compensation are precise, and contract terms are being met.

But data assembly presents a large hurdle. Here's what makes it challenging, how it can be implemented effectively, and the significant insights and savings it can provide.

The Constraints of Siloed Information

Solar monitoring portals may reliably report electricity production, often in 15-minute intervals. But they're limited in three important ways:

  • Their scope is narrow. An organization's portfolio might comprise numerous systems, but each portal sees only its own system's equipment.
  • They are shallow. Most show production and little else. Expected production, when available, varies in quality.
  • Their financial figures can be misleading. Savings and avoided emissions are often calculated using a flat cost per kilowatt-hour. Actual electricity costs vary by time, season, demand charges, and rate structure, so these estimates can be materially inaccurate.

While a monitoring portal can show the amount of electricity a system has generated, it can't interpret the significance of that number. And it won't answer portfolio-level or financial questions, either.

The information needed to answer those questions lives in many different places:

  • Production data
  • Design specifications
  • Utility bills
  • Tariff schedules
  • Contract terms
  • Weather data
  • Interval consumption data

Each of these pieces is individually available. But none of them talk to each other.

The work is in making the connections: matching a vendor system identifier to an asset record, that asset to the meters it feeds, those meters to the accounts that get billed, and all of it to a weather series and a contract. Only then can organizations evaluate production against actual site performance and cost.

From Data Assembly to Actionable Analytics

Bringing the data together requires more than simply feeding it into a single system, though. It's an exercise in reconciliation and contextualization.

Combining the pieces exposes disagreements that were previously hidden. For instance, data that looked clean in isolation may turn out to have gaps, scaling changes, sensor faults, or duplicate records. Done well, the reconciliation process can tease out data problems from actual equipment problems.

The contract layer adds another dimension. Commercial solar is often governed by a power purchase agreement, with terms that determine the financial outcome. Those terms typically live in PDFs rather than databases, as do important tariff details. Document extraction now makes it possible to bring those terms into the data model alongside production and weather.

This information changes the questions an organization can answer. Instead of treating contract review as a one-time exercise, it can continuously ask: Is the system meeting its production guarantee? How does the contracted rate compare with current electricity costs? What is underperformance costing? Is a buyout worth evaluating?

Weather normalization then puts performance into context by establishing what each system should have produced under actual conditions. Two baselines reveal different problems: comparison with recent performance detects changes like outages or equipment failures, while comparison with the original design estimate reveals chronic underperformance.

Normalizing production by system size makes different systems comparable, too.

In addition, applying the actual tariff (not a flat electricity rate) turns production data into a meaningful financial picture, including time-of-use pricing, demand charges, and net-metering credits.

AI has an important role in making this process practical, particularly in extracting information from contracts, bills, and construction documents and identifying patterns across large datasets. But it should not replace the performance baseline itself, because a model that learns a system's historical behavior can also learn its degradation or faults and mistake poor performance for normal performance. For verification, the baseline needs to be interpretable, grounded in the physics of the system, and defensible when financial guarantees are at stake.

Unlocking the Big Picture Drives Results

Providing our Empower platform clients with this robust, centralized data perspective has helped them to uncover and rectify an array of problems with their solar and storage systems. For example:

  • A rooftop array had been producing roughly 40 percent below its design estimate for as long as records existed. Every monitoring platform reported it as operating normally. Nothing had failed, and nothing had changed. But it had never delivered what it was sold as delivering.
  • A system had lost roughly half its output over five months, while the platform reported data continuously throughout. Since production was still happening (just less of it), the client remained unaware of the underperformance.
  • A client was being billed for twice as many net-metered accounts as their own records described, including a credit sharing arrangement across a dozen tenant meters that appeared in no documentation.
  • An apparent 25 percent production loss turned out to be a sensor recalibration.
  • A client was being undercompensated as per their power purchase agreement due to lack of understanding of the contract terms and how to evaluate the economics of the systems.
  • A client's large solar array went offline for six months. Without weather-normalized predictive analytics, they couldn't evaluate the value of the lost production relative to grid-purchased electricity to qualify for cost recovery from their solar provider.

Solar and storage portfolios generate plenty of data. The challenge is transforming that fragmented information into a comprehensive, living picture of their performance and value that supports budgeting, informed decision-making, and strategic planning.

Ready to learn more about how data assembly can benefit your existing or future solar program?

Drop Us a Line
Picture of Julia Soplop and Jeff Soplop

Julia Soplop and Jeff Soplop