ARR 2019 in Practice: How New Rainfall Data Changes OSD Design

If you have been building in Sydney for more than a decade, you have probably noticed that stormwater reports look different these days. Where an OSD calculation once ran off a single design storm pulled from a 1987 rainfall chart, your engineer now talks about ensembles, temporal patterns, pre-burst rainfall and Data Hub losses. That is not consultant jargon for the sake of it. It is Australian Rainfall and Runoff 2019 — the national guideline that replaced ARR 1987 — and it has genuinely changed how on-site detention and drainage systems are sized across NSW.

For builders, developers and private certifiers, the details of hydrology are someone else’s problem right up until they are not. An OSD tank sized under the wrong method can trigger a council request for information, a redesign mid-construction, or a compliance certificate that a certifier cannot sign off. This article explains what actually changed under ARR 2019, why it matters for detention sizing in Greater Sydney, and what you should be asking your stormwater engineer before the slab goes down.

Why the Single Design Storm Is Obsolete

Under ARR 1987, the process was straightforward. You picked a storm frequency — say the old 100-year event — read an intensity off a chart for a range of durations, applied one standard temporal pattern per duration, and ran the calculation. One storm in, one answer out. Every engineer using the same inputs got the same result, which made checking easy but concealed a fundamental problem: real storms do not follow a single standard shape.

Two storms can deliver identical total rainfall over 90 minutes and yet produce very different peak flows, depending on whether the intense burst arrives early, late or in the middle. A front-loaded storm might fill a detention tank before the peak arrives; a back-loaded one might catch it near full. ARR 1987’s single pattern averaged this variability away, and decades of research showed the results could be biased — sometimes conservative, sometimes not, with no way of knowing which without doing the analysis properly.

ARR 2019 fixed this by replacing the single design storm with an ensemble approach, underpinned by far better rainfall data.

The Big Changes Under ARR 2019

New 2016 IFD data

The intensity-frequency-duration (IFD) data behind every design storm was regenerated by the Bureau of Meteorology in 2016 using a much longer and denser rainfall record than was available in the 1980s. For some Sydney locations the new IFDs are higher than the old charts; for others they are lower. There is no universal rule — the change is location-specific, which is one reason legacy OSD calculations cannot simply be assumed conservative.

An ensemble of ten temporal patterns

This is the change with the biggest practical impact. Instead of one standard storm shape per duration, ARR 2019 provides ten temporal patterns for each duration, downloaded from the ARR Data Hub for the relevant region. The engineer runs all ten through the model and examines the spread of results. Some patterns are front-loaded, some back-loaded, some double-peaked — reflecting the genuine variability of real storms.

The consequence is that a design flow or detention volume is no longer a single deterministic number. It is a statistic drawn from an ensemble. Practice varies on which statistic to adopt: many Sydney councils accept the mean or median of the ensemble peaks for OSD sizing, and some are comfortable with DRAINS reporting the average of ensembles. Others have their own stated preference in their engineering specifications. A competent stormwater engineer confirms the council’s position before modelling begins, not after the RFI arrives.

Critical duration means something different now

Under the old method, finding the critical duration meant running a handful of single storms and picking the one that produced the highest peak. Under ARR 2019, the critical duration must be found by testing the full ensemble across the range of durations — for a typical infill or medium-density site that can mean dozens of model runs per AEP. And here is the part that catches people out: the critical duration for peak flow leaving an undeveloped site is often not the critical duration for storage in a detention tank. Short, intense bursts govern pipe and pit sizing; longer storms frequently govern OSD volume because they deliver more total water while the outlet is throttled. The ensemble approach makes this distinction explicit rather than leaving it to a rule of thumb.

Pre-burst rainfall and losses

ARR 2019 also recognises that design storm data represents a rainfall burst — the intense core of a storm — and that real storms usually include rain before that burst. This pre-burst rainfall wets the catchment, so by the time the design burst arrives, less rainfall soaks into the ground and more runs off. The ARR Data Hub provides pre-burst depths along with initial and continuing loss values for each location, and these feed directly into the model. Get the losses wrong and both peak flows and detention volumes shift, particularly on sites with significant pervious area.

Embedded bursts

One quirk worth knowing about: some of the ten temporal patterns contain embedded bursts — short periods within the pattern that are statistically rarer than the burst as a whole. A pattern representing a 1% AEP storm over three hours might contain a 30-minute window that is rarer than 1% AEP. Left unchecked, these can distort results, especially for small, fast-responding urban catchments — which is exactly what most development sites are. Experienced modellers review the patterns and treat embedded bursts appropriately in line with current guidance. It is a detail, but it is the kind of detail that separates a defensible stormwater report from one that unravels under council review.

What This Means for OSD Sizing

On-site stormwater detention in NSW revolves around two numbers: the permissible site discharge (PSD), which caps the rate at which your site may release stormwater, and the site storage requirement (SSR), the volume you must detain to stay under that cap. In the Parramatta and Hills districts, many councils still reference the UPRCT On-Site Stormwater Detention Handbook, developed by the Upper Parramatta River Catchment Trust, which prescribes PSD and SSR values by catchment. Other councils set their own rates or require the post-development discharge to match pre-development flows across a range of storm frequencies.

ARR 2019 changes the arithmetic underneath all of this wherever the modelling is done from first principles. Because the 2016 IFDs differ from the 1987 charts, and because ensemble statistics replace single-storm peaks, the pre-development flows, post-development flows and the storage needed to bridge them can all shift. On some sites the required OSD volume grows compared with a legacy calculation; on others it shrinks. The point is not that ARR 2019 makes tanks universally bigger — it is that an OSD design carried over from an old report, a neighbouring project or a previous DA cannot be assumed valid. It needs to be re-run with current data.

For builders, the practical risk is programme and cost. An underground OSD tank is one of the first things constructed and one of the most expensive to change. Discovering at construction certificate stage that the tank on the architect’s plans was sized on superseded rainfall data is an avoidable headache.

How Councils Are Handling the Transition

Unevenly, is the honest answer. NSW councils adopted ARR 2019 at different speeds and in different ways. Some updated their DCPs and engineering specifications promptly and now expect full ensemble modelling with Data Hub inputs. Others still reference deemed-to-comply PSD/SSR rates derived under older methods — the UPRCT handbook being the most prominent example, still referenced by several north-western Sydney councils — while expecting ARR 2019 hydrology for anything modelled from scratch. A few sit somewhere in between, accepting either approach provided the engineer states the method clearly.

This patchwork means the same dual occupancy design could face different stormwater requirements in adjoining LGAs. There is no substitute for reading the specific council’s DCP and engineering standards at feasibility stage. If your project spans a boundary or you build across multiple council areas, do not assume the last approval sets the template for the next one.

The Practical Workflow in DRAINS

DRAINS remains the dominant modelling package for site drainage and OSD design in NSW, and it handles the ARR 2019 workflow natively. A typical project sequence looks like this:

• Download site-specific 2016 IFD data from the Bureau of Meteorology and the ensemble temporal patterns, pre-burst depths and loss values from the ARR Data Hub.
• Build the catchment model — pervious and impervious areas, flow paths, pits, pipes and the proposed OSD arrangement with its orifice or throttle.
• Run the full ensemble across a range of durations for each AEP the council requires, identifying the critical duration for both peak flow and storage separately.
• Extract the council’s preferred statistic — mean or median of the ensemble peaks, or the average-of-ensembles output — and verify PSD compliance and SSR adequacy.
• Check the high-early-discharge arrangement, overflow provisions and freeboard, then document the method, inputs and results in a report a council engineer can audit.

None of this is exotic for a practice that does it every week. But it is substantially more work than the 1987-era single-storm calculation, and reports that skip steps — a single temporal pattern, no pre-burst, losses assumed rather than sourced — are increasingly picked up by council development engineers and returned with an RFI.

What Builders and Developers Should Ask Their Engineer

You do not need to understand hydrology to procure it well. A few direct questions will tell you quickly whether your stormwater consultant is working to current standards:

• Are you designing to ARR 2019 with the 2016 IFDs and Data Hub inputs? The answer should be an unqualified yes, with the site-specific data sourced rather than recycled.
• Have you confirmed this council’s OSD method and ensemble preference? The engineer should be able to name the DCP clause or engineering specification they are designing to, and whether the council follows the UPRCT handbook, its own PSD/SSR rates, or pre- versus post-development matching.
• What governs the tank size — and what is the critical duration for storage? A modeller who has actually run the ensemble will answer this without hesitation.
• Will this report survive certifier and council review? Ask whether the methodology, inputs and ensemble results are documented well enough for a third party to check.

Ask these at engagement, before the architect locks in a basement footprint or a landscape plan that leaves nowhere for detention. Stormwater is cheap to solve on paper and expensive to solve in concrete.

Frequently Asked Questions

Does ARR 2019 mean my OSD tank has to be bigger?

Not necessarily. The 2016 IFD data is higher than the old 1987 charts in some Sydney locations and lower in others, and the ensemble method changes how peaks are calculated. Some sites end up with more storage, some with less. The only way to know is to run the current data for your specific site and council requirements — which is exactly why old reports should not be recycled.

Can I reuse a stormwater report prepared under ARR 1987?

Generally no. Most NSW councils now expect ARR 2019 hydrology for new development applications, and a report based on superseded rainfall data invites an RFI or refusal. If the council applies a deemed-to-comply method such as the UPRCT handbook rates, parts of the approach may carry over, but the report itself should be prepared fresh against the current DCP and engineering standards.

What is the ARR Data Hub and why does it matter?

The ARR Data Hub is the online repository that supplies the location-specific inputs ARR 2019 requires — the ten ensemble temporal patterns per duration, pre-burst rainfall depths, and initial and continuing loss values. These inputs, combined with the Bureau of Meteorology’s 2016 IFD data, are what make an ARR 2019 model site-specific rather than generic.

Do all Sydney councils apply ARR 2019 the same way?

No. Councils vary in which ensemble statistic they accept for OSD sizing, which AEPs they require you to test, and whether they still apply prescriptive PSD/SSR rates such as those in the UPRCT handbook. Always check the specific council’s DCP and engineering specification — or engage a stormwater engineer who already knows them.

Need an OSD or drainage design that meets current ARR 2019 requirements? Contrive Consultants prepares council-ready stormwater and detention designs across Rouse Hill, the Hills District and Greater Sydney, with full ensemble modelling in DRAINS tailored to your council’s standards. Explore our stormwater engineering services or get in touch for a fee proposal — call +61 497 848 111 or email info@contriveconsultants.com.au.