Post: Adding a battery to a solar farm doesn’t create a hybrid. Optimising them together does

Adding a battery to a solar farm doesn’t create a hybrid. Optimising them together does

Guest post by GridBeyond Business Development Director (AU), Mark Netto

Most battery optimisation platforms were built for standalone BESS and later adapted for hybrids. That distinction matters — particularly in SA and QLD, where negative prices now affect around 30% of intervals and the evening peak window is where most revenue is made or missed.

DC-coupled hybrid assets have specific optimisation requirements — DC charging efficiency, loss factor accounting, ramp rate scope — that adapted platforms often don’t handle correctly. Our analysis suggests correctly modelling DC charging paths alone can contribute 5–10% of additional project returns.

Before selecting a platform for a hybrid project, it’s worth understanding whether you’re buying something designed for this asset class or something adapted from somewhere else.

A growing proportion of the sub-5MW pipeline is no longer standalone battery storage.

Co-located solar and storage. Retrofitted solar farms adding BESS. DC-coupled systems. Behind-the-meter hybrids. The configurations vary, but the underlying trend is consistent — more non-scheduled hybrid projects are being built, and that number is accelerating.

This creates an optimisation problem that is meaningfully different from standalone BESS. Not more complicated in ways that are hard to solve — but different in ways that many platforms were not originally designed for.

That distinction is starting to matter.

The standalone assumption

Most battery optimisation platforms were built for single-technology BESS and later adapted to handle hybrid assets.

That is a reasonable product evolution. But it leaves a residue in how the optimisation logic works.

A standalone battery has one job: dispatch at the right time. The platform forecasts prices, manages state of charge, submits FCAS offers, and sequences charge and discharge cycles to maximise revenue. The inputs are relatively clean — grid connection, AEMO price signal, telemetry from the battery.

Add solar generation behind the same connection point, and the problem changes.

The platform now needs to account for PV output in real time. It needs to decide, every five minutes, whether available solar generation should be exported to the grid, used to charge the battery, or some combination of both — and that decision needs to be made against a forecast of what prices are going to do over the next several hours.

These are not two separate decisions that can be made sequentially. They are one decision, and they need to be made together.

The negative price problem

This is where the gap between genuine hybrid platforms and adapted standalone optimisers becomes most visible.

In South Australia, approximately 30% of all intervals now see near-zero or negative spot prices, driven by excess rooftop and utility solar generation during daylight hours.* In Queensland, that proportion is rising.

For a standalone battery, negative prices are primarily a charging opportunity. The logic is relatively straightforward.

For a hybrid asset, the calculation is fundamentally different.

During a negative price interval:

• the solar farm is generating
• exporting that generation incurs a cost
• the battery may or may not have capacity to absorb it
• the connection point has physical export limits
• solar clipping may result in lost generation if not actively managed

Somewhere in the next few hours, the evening peak arrives — the point at which supply tightens and prices can spike above $5,000/MWh. That is when the battery needs to be positioned to discharge. Getting the state of charge right for that window is the central optimisation challenge of a hybrid asset.

Getting it right requires a platform that is simultaneously managing:

• PV generation forecasts
• battery state of charge
• connection point constraints
• forward price expectations

—and revising all of those every five minutes.

A platform that handles solar and battery as separate modules is not doing this. It is making two sequential decisions and hoping they are compatible.

Where adapted platforms leave money on the table

Working with DC-coupled hybrid assets at the project level, a few specific failure modes become apparent. They are worth naming because they are not obvious from the outside — but they are structural, and they compound over time.

DC charging efficiency

In a DC-coupled system, charging the battery directly from the solar string skips the inverter stage entirely. The efficiency difference is significant — DC solar charging can run at over 98% versus a full AC round-trip at closer to 82%.

An optimiser that models all charging paths as equivalent will consistently undervalue solar charging relative to grid charging.

During negative price periods, when the choice between the two is most consequential, this leads to systematically suboptimal decisions.

Our analysis indicates that correctly modelling DC charging paths can contribute an additional 5–10% of project returns over the asset’s life — and that requires forecasting DC output directly, not inferring it from AC metering.

Loss factor accounting

A DC-coupled hybrid asset carries both a transmission loss factor and a distribution loss factor, and a platform that does not correctly model these will misprice every charge and discharge decision.

In any given interval the error is small. Across a full year of dispatch it compounds into something material.

Ramp rate scope

DNSP connection agreements typically impose a ramp rate constraint on the asset. In many cases, however, that constraint applies to energy dispatch and not to FCAS.

An optimiser that applies the ramp limit uniformly will unnecessarily restrict availability offers, leaving contingency raise and lower revenue uncaptured.

It is a quiet, consistent drag on performance that is invisible unless you are looking for it.

None of these are fringe edge cases. They are the normal operating parameters of a DC-coupled hybrid project.

A platform built for this asset class handles them as standard. A platform adapted from standalone BESS may not — and the only way to know is to ask directly.

What to ask any provider

If you are evaluating platforms for a hybrid project, a few questions are worth putting to any provider:

• Is your hybrid optimisation a single co-optimised solve, or do solar and battery dispatch run as separate modules?
• Do you forecast DC output directly, or do you rely on AC metering to infer solar generation?
• How does your platform model the efficiency difference between DC solar charging and AC grid charging?
• Does your ramp rate constraint logic distinguish between energy and FCAS markets?

The answers will quickly reveal whether you are dealing with a platform designed for hybrid assets or one adapted from somewhere else.

Why it matters now

The volume of hybrid projects entering the sub-5MW market is increasing.

The market conditions those projects face — particularly in SA and QLD — are becoming more complex. Negative price frequency is rising. The evening peak windows where revenue is concentrated are becoming shorter and harder to predict.

In that environment, the gap between a platform that treats solar and storage as one asset and one that treats them as two is going to widen.

For developers building hybrid portfolios, this is worth understanding before selecting a platform — not after the first year of underperformance.

* Source: AEMO Quarterly Energy Dynamics

Series note

This is the third in a series on what separates high-performing sub-5MW battery portfolios. Previous articles can be found here and here.

Next article: Operational resilience and future-proofing — why silent failures, telemetry gaps and evolving NEM requirements are becoming one of the biggest risks in sub-5MW portfolios