Microgrids, VPPs and Dispatch Algorithms: Where the R&D Actually Sits in an Energy Software Project

As at 27 July 2026, the Australian Government had announced reforms to the R&D Tax Incentive in the 2026–27 Federal Budget, intended to apply to income years starting on or after 1 July 2028. Until those changes take effect, the program continues to operate under the current legislation, and we cover the announced measures separately in the proposed $50,000 Budget measure.

You have forty-something behind-the-meter batteries, three embedded networks, a bidding stack, a forecasting service and a control plane tying it together. Nothing quite like it exists in Australia. So when the accountant asks what to register for the R&D Tax Incentive, the first instinct is "all of it".

That instinct can create eligibility risk for a specific reason. The R&DTI does not assess projects. business.gov.au puts it plainly: "under the R&DTI, eligibility is determined at the activity level, not the project level" (check if you are eligible). A distributed-energy business is a dense mix of activities — some experimental, most operational — running on the same hardware at the same time.

We are an RSP, not a registered tax agent. What we can help with is the thing this turns on: a field trial genuinely separable from running the business.

The Conclusion First: Assess the Experiment, Not the Fleet as a Whole

Section 355-25(1) of the Income Tax Assessment Act 1997 defines core R&D activities as experimental activities whose outcome "cannot be known or determined in advance ... but can only be determined by applying a systematic progression of work" that is "based on principles of established science" and "proceeds from hypothesis to experiment, observation and evaluation, and leads to logical conclusions", conducted for the purpose of generating new knowledge (ITAA 1997).

Run a typical DER program through that and the buckets separate. For the internal-administration exclusion, assess each software module according to its dominant purpose of use, not simply what the module is called.

That software split can be easy to misclassify. Section 355-25(2)(h) is not a rule about control, trading or dispatch software as a category: it excludes software developed, modified or customised "for the dominant purpose of use by" the developer, an entity connected with the developer, or an affiliate (or an entity of which the developer is an affiliate) "for their internal administration (including the internal administration of their business functions)". business.gov.au states the corollary on its excluded R&D activities page: the exclusion "does not apply" where the dominant purpose of use is "other than for internal business administration", and it lists among software outside the limb software that "is used by your customers and is unrelated to your internal administration". However, a small number of external licences does not by itself determine the software's dominant purpose of use.

So a dispatch engine that delivers energy services to external customers, or bids a portfolio into a market, is not internal administration merely because your own operators also sit in front of it. The analysis runs module by module; where one module genuinely serves both, you weigh the competing purposes, as set out below. Nor is exclusion from core the end of the story: an excluded activity may still be assessed as supporting if directly related to a core activity and conducted for the dominant purpose of supporting it (more in our software and AI material).

Why Simulation May Not Determine the Outcome

The unknown here is not whether the code compiles. It may be how a heterogeneous fleet of inverters, batteries and loads responds when a new control policy meets real network, weather and operating conditions. Limitations in available simulation models may support the assessment, but they do not determine eligibility on their own. The relevant question remains whether a competent professional could have known or determined the outcome in advance without conducting the experiment:

Building a dispatch optimiser using established methods may be routine engineering rather than core R&D where its outcome can be determined in advance.

Designing a Field Trial That Is Separable from Ordinary Operation

The structural problem, stated exactly: the assets you experiment on are the assets that earn your revenue. The battery testing your hypothesis on Tuesday is bidding into the market on Wednesday. Without a clear distinction between the two, it may be difficult to identify and substantiate the experimental activity.

What Division 355 requires is an experimental activity with an unknown outcome, determined by a systematic progression of work that proceeds "from hypothesis to experiment, observation and evaluation, and leads to logical conclusions" (s 355-25(1)(a)(ii)). On form, business.gov.au adds only that experiments "can take many different forms" but "must aim to test a hypothesis" (conducting core R&D activities).

With that boundary clear, here is the design we use when trial assets are also trading. Each element is most useful settled before the trial starts:

Then the element founders skip: where the trial ends. Once the hypothesis has been evaluated, that experimental activity is generally complete unless a further unresolved technical question gives rise to a new hypothesis and systematic progression of work. Rolling the winning controller across the fleet is deployment; extending the window because the results are commercially useful does not extend the core activity.

The Dominant-Purpose Trap When the Trial Earns Revenue

Supporting R&D activities are "activities directly related to core R&D activities" (s 355-30(1)). Section 355-30(2) adds a harder test: if an activity is on the excluded list, or produces goods or services, or is directly related to producing them, it is supporting only if undertaken for the dominant purpose of supporting core R&D activities.

Supporting activities conducted as part of a field trial on revenue-earning batteries may produce, or be directly related to producing, services. Where they do, they qualify as supporting R&D activities only if undertaken for the dominant purpose of supporting core R&D activities. business.gov.au's conducting supporting R&D activities page applies it "regardless of the terms of the services", including "services your business provides to itself", and gives the line that decides most DER cases: activities are unlikely to have the dominant purpose of supporting core R&D where they "produce a good or service that you do not use to conduct an experiment" or that "you would undertake if you did not conduct R&D".

Applied honestly, ordinary operation of your fleet is usually not supporting, because you would do it anyway; the candidates are the incremental activities that exist only because the trial does — trial-only instrumentation, extra data capture, extra site visits, control-cohort management.

Weighing competing purposes inside one platform: For the harder case — a module serving both your own administration and your customers — business.gov.au sets out a sequence: first assess whether the software is for use by you, an entity connected with you or an affiliate; if so, assess whether the dominant purpose of that use is internal administration, or the internal administration of business functions. In deciding, you weigh which entities the software is being developed for, "how those entities will use that software", and "the strength of each purpose and how important each is compared to the other purposes" — and, critically, "There may be several purposes for using software, but you can only have one dominant purpose."

So if a settlement engine or ops console exists predominantly to run your business, the limb is a live risk however hard the engineering, and a few external licences will not fix it. Show your working (what does not qualify).

Worked Example: An Adelaide VPP Operator (Illustrative Only)

A South Australian aggregator runs 900 residential batteries and three embedded-network microgrids. It wants to know whether a reserve-aware dispatch policy can hold islanding capability at the microgrid sites without surrendering market availability. Figures are illustrative, on assumed facts, not a projection for your business.

Suppose that scopes to $180,000 of candidate eligible R&D expenditure for the income year. Which offset applies, and whether it is refundable, is a separate question: it turns on eligibility, aggregated turnover and whether the company is controlled by one or more income-tax-exempt entities, not on whether the company is in a loss position. What comes back as cash also depends on its tax liabilities. See refundable vs non-refundable offset; the calculation belongs to your registered tax agent.

Where an RSP Fits, and the $20,000 Threshold

The deliverable that makes this claim defensible is not a tax document but a trial protocol — hypothesis, comparator, measurement plan, evaluation — designed to run on assets that are simultaneously trading. A Registered Research Service Provider may assist with planning and conducting this research work. Ignition Research is RSP000047, based at Lot Fourteen in Adelaide (Adelaide page).

It matters most at the small end, where the experimental slice of a first trial is often modest. Per business.gov.au's get help from a research service provider page, RSP-conducted eligible R&D activities can be claimed even where the usual $20,000 R&D expenditure threshold is not met — the mechanism is s 355-100(2) of the ITAA 1997, and it is narrow. Where total notional deductions are below A$20,000, the offset base is generally limited to qualifying expenditure incurred to a non-associate RSP for services within a field for which it is registered, together with eligible CRC Program contributions. Other in-house amounts do not automatically form part of that below-threshold offset base. And using an RSP does not guarantee eligibility — you still self-assess. More in claiming R&D under $20,000.

One more point concerns grant-funded DER trials. Where ARENA or a state program provides funding that relates to expenditure taken into account in calculating the R&D tax offset, a clawback adjustment may apply. Under Subdivision 355-G an amount arises where an entity "receives or becomes entitled to receive a recoupment" from an Australian government agency or a state body in relation to expenditure for which it has obtained R&DTI offsets (s 355-440), and that amount is included in the entity's assessable income (s 355-450) — as the ATO puts it, clawback "doesn't decrease the grant or offset you receive. Instead, it increases your assessable income" (ATO). Quantifying it belongs to your registered tax agent.

The 2026-27 Budget announced proposed R&DTI changes for income years starting on or after 1 July 2028. They are not current law; see our dedicated Budget update for the proposed measures and their status.

Frequently Asked Questions

Q: Can microgrid companies claim the R&D Tax Incentive?
A: A microgrid company can register eligible R&D activities, but eligibility is assessed at the activity level, not the project level. Building and operating a microgrid is generally commercial engineering and operation; what may be assessable as core R&D is a specific experiment inside it — a control or dispatch policy whose real-world behaviour cannot be determined in advance and tested through a systematic progression of work. You self-assess.

Q: Is energy management software eligible for the R&D Tax Incentive?
A: It depends on each module and, above all, on the dominant purpose of its use. Section 355-25(2)(h) excludes from core R&D software developed for the dominant purpose of use by you, an entity connected with you, or an affiliate, for internal administration. Software whose dominant purpose of use is delivering energy services to customers or to a market is generally not caught — business.gov.au says the exclusion "does not apply" where that purpose is other than internal business administration. It must still meet the core test, or qualify as supporting.

Q: Is a battery or VPP field trial an eligible R&D activity?
A: It may be. The statutory test is in s 355-25(1): an unknown outcome resolved by a systematic progression of work from hypothesis to experiment, observation and evaluation, leading to logical conclusions, for the purpose of generating new knowledge. Division 355 does not require a control group, a fixed window or a pre-written measurement plan — those are evidence practices, and the most practical way we know to show the activity was an experiment. Because trial assets also earn revenue, activities producing goods or services qualify as supporting only where undertaken for the dominant purpose of supporting a core R&D activity (s 355-30(2)).

Q: Is commissioning and operating a microgrid an R&D activity?
A: Commissioning, inverter and protection configuration and ongoing operation are generally not core R&D where their outcomes can be determined in advance. Testing associated with demonstrating compliance with statutory requirements or standards is excluded from core R&D under s 355-25(2)(f). Directly related activities may qualify as supporting R&D where the applicable requirements, including the dominant-purpose test where triggered, are met. Confirm your position with your registered tax agent.

Sources & Further Reading

• business.gov.au — Check if you are eligible for the R&D Tax Incentive

• business.gov.au — Conducting core R&D activities

• business.gov.au — Conducting supporting R&D activities

• business.gov.au — Excluded R&D activities

• business.gov.au — Get help from a research service provider

• legislation.gov.au — ITAA 1997, Division 355 (ss 355-25, 355-30, 355-100, 355-440 to 355-450)

• ato.gov.au — Clawback of R&D tax incentive offset and catch-up deductions

• Related: renewable energy · what does not qualify · under $20,000 · software and AI · refundable vs non-refundable