AS 2870 site classification explained: what classes A to P mean

An AS 2870 site classification (often written AS2870) is a single letter that tells the footing designer how much the ground under a house is expected to move as the clay in it wets and dries. Class A is sand or rock that barely moves; S, M, H1, H2 and E rank reactive clay from slight to extreme; P marks a problem site that needs its own engineering. The certifier expects to see the letter before the slab is designed.

The standard exists to replace inconsistent local practice. Before 1986, reactive clay was handled differently from state to state and even council to council. Houses cracked, owners went to court over quite minor damage, and engineers appearing as expert witnesses discredited each other over how the clay should have been identified (Cameron 2018). AS 2870 gave the country one method. The current edition is AS 2870-2011.

What do the AS 2870 site classification letters mean?

Each reactive class is a band of predicted movement. The quantity is the characteristic surface movement, written ys: how far the ground surface would rise and fall between its driest and wettest design states with no house on it.

Class	Predicted surface movement (ys)	Ground
A	little or none	mostly sand and rock
S	up to 20 mm	slightly reactive clay
M	over 20 to 40 mm	moderately reactive clay
H1	over 40 to 60 mm	highly reactive clay
H2	over 60 to 75 mm	highly reactive clay, very high movement
E	over 75 mm	extremely reactive clay
P	not a movement band	a problem site: fill, soft ground, abnormal moisture, instability

ys is reported to the nearest 5 mm, so the estimate is coarser than the band edges suggest. Where seasonal drying reaches 3 m or deeper, as in Adelaide, the class carries a "-D" suffix for deep-seated movement, for example H1-D.

Each step up means a stiffer, deeper, more expensive slab. The standard's off-the-shelf (deemed-to-comply) designs are sized for the top of each band, and they stop at H2 (Cameron 2018). A site at 45 mm gets the same H1 slab as one at 60 mm, and a Class E site gets no standard design at all. What each class means for the slab covers the footing side, and Class P sites the problem class.

How is the letter worked out?

Three inputs go into ys, and they are the terms you will see on a report.

How reactive the clay is. Measured in the lab as the shrink-swell index, Iss: how much a sample changes length for each unit change in soil suction (the force with which the soil holds water, measured on a log scale in pF). Two other tests, loaded shrinkage and core shrinkage, are also accepted, but the shrink-swell test is generally preferred because it captures both swelling and shrinking (Cameron 2018).

How deep the seasonal wetting and drying reaches. AS 2870 calls this Hs, the depth of design suction change. It is set by climate, not by the clay. The standard lists it for the main cities: Sydney 1.5 to 1.8 m, Brisbane and Ipswich 1.5 to 2.3 m, Melbourne 1.8 to 2.3 m, Perth 1.8 m and Adelaide 4.0 m, with the lower end for wetter coastal and hill suburbs. Elsewhere it comes from the Thornthwaite Moisture Index (TMI), a wetness score built from rainfall and evaporation, which puts a site in one of six zones, with Hs of 1.5 m in the wettest and over 4 m in arid country.

How far the surface soil swings. A design suction change at the surface, 1.2 pF for every listed city.

The calculation adds up the movement of each clay layer down to Hs. The top part of the profile, the cracked zone where shrinkage cracks open in dry weather, is allowed to move more freely: 0.75Hs in Adelaide and Melbourne and 0.5Hs elsewhere, unless better information is available.

Change an input and the letter can change with nothing in the ground being different. Bruce Hargreaves, looking back on twenty years of AS 2870 practice in South East Queensland, worked through what happened when the 1996 edition deepened Hs and halved the cracked zone. In his example, an Ipswich site that calculated at 65 mm under the old parameters would come out at 115 mm under the new ones, with the same method and the same clay (Hargreaves 2005). His point was not that the old classifications were wrong, but that the letter is only as good as its inputs. Ask which Hs a report used.

The other thing to ask is whether anyone tested the clay. Often nobody did. Very little reactivity testing has been done in South Australia since 1986, because classifiers there judge the local clays by look and feel (Cameron 2018), and SEQ practice has leaned heavily on experience too (Hargreaves 2005). The standard allows a qualified engineer or engineering geologist to judge reactivity by look and feel, but the 2011 edition makes them check that judgement against lab tests at least every six months and once in every 50 sites.

If you hold lab Iss results, the free AS 2870 site classification calculator runs the full calculation with a layer-by-layer breakdown.

Why does the same clay get a different letter in different cities?

Because Hs is climatic. Put the same clay in a profile that dries to 4 m instead of 1.5 m and far more of it moves. Adelaide's Keswick and Hindmarsh clays are generally less reactive than Melbourne's basalt clays, yet they cause more trouble, largely because of that 4 m drying depth (Cameron 2018). Why the same clay classifies differently compares Sydney, Melbourne and Adelaide in detail, and our SEQ guide covers Brisbane, Ipswich and Toowoomba.

Two cautions. TMI maps by different authors, such as Chan and Mostyn (2009) for NSW, use different methods and do not always agree where they meet, so the zone is an engineer's judgement, not a lookup. And the climate moves: Lopes and Osman (2010) found Victoria has been getting drier since the late 1940s, and that a rise of about 20 percent in predicted movement is enough to lift a borderline site a class.

How do you get an AS 2870 site classification?

A site investigation. A geotechnical engineer or technician drills or digs the lot, logs the profile, and tests the clay or assigns its reactivity from local experience. AS 2870 asks for at least one borehole or pit per building site, and at least three where Hs is 3 m or more or the profile is known to vary. Holes go to at least 0.75Hs, with a 1.5 m minimum unless rock is hit first. This is the only route to a classification a certifier will accept.

A desktop study is not one. Mapped geology, soil surveys, nearby bore logs and the climate zone can set out the reactivity range to expect before anyone drills, which is useful before purchase and for briefing the investigation. It cannot give the letter. Li et al. (2016) tested clay at 47 Melbourne sites and found a wide spread of results within the same mapped geological unit. Only a hole on your lot tells you where in that spread you sit. The site investigation decides. What a desktop study covers sets out the boundary.

Why does getting the letter right matter?

An under-classified site gets a slab designed for less movement than it will see, and the result is the familiar pattern of cracked brickwork, sticking doors and gapped cornices that reactive soil damage describes. An over-classified site just costs more in every footing.

Fill is one place where under-classification starts. The standard gives controlled clay fill placed less than five years before building no cracked zone, so it is assumed to move more than the same clay left in place. In Cameron's (2018) worked example, 0.8 m of reactive fill took a site from H1 (58.5 mm) to E (89 mm). Hargreaves (2005) found increases of 50 to 66 percent in his SEQ examples, and warned that the classifier may report on the natural clay while the slab designer is never told the fill above it can heave.

The standard's fill rules (clause 2.5.3) reflect this. Controlled fill, placed and tested to AS 3798, up to 0.4 m deep (0.8 m for sand) leaves the site classified as the natural ground was before filling. Deeper controlled fill other than sand makes the site Class P unless an engineer reclassifies it from the expected movement of the fill and the ground beneath. Uncontrolled fill makes the site Class P; where it is shallow, that can be avoided only by taking every footing through it to natural soil.

Common questions

What does an AS 2870 site classification mean? A is non-reactive, S slightly, M moderately, H1 and H2 highly (H2 with very high movement) and E extremely reactive clay, banded by predicted surface movement (S to 20 mm, M 20 to 40 mm, H1 40 to 60 mm, H2 60 to 75 mm, E over 75 mm). P means a problem site that needs specific engineering.

Can a site classification be done without a soil test? A compliant classification for footing design needs a field investigation. A desktop assessment can set out the reactivity range to expect from geology, soil mapping, nearby bores and the climate zone, which is useful before purchase and for scoping the investigation, but it is not a classification.

Why is Brisbane's classification different from Melbourne's for similar clay? The depth of seasonal moisture change (Hs) is set by climate: AS 2870-2011 gives 1.5 to 2.3 m for Brisbane and Ipswich and 1.8 to 2.3 m for Melbourne, and Melbourne also uses a deeper cracked zone. Deeper moisture change means more of the clay profile contributing to movement, and a higher ys for the same soil.

---

Before you commission drilling, pull the desktop picture for the lot: mapped geology, soils, nearby bore logs, groundwater and the reactivity range to expect, in one report. Generate a desktop report for your address or see what a sample report contains.

LayeredGeo compiles geology, soils, groundwater and site data into an automated geotechnical desktop report for any address in Queensland, New South Wales, Victoria and South Australia.

Sources

• Cameron, D. A. (2018). Dealing with reactive clay soils through a national standard. Australian Geomechanics, Vol 53 No 1. geomechanics.org.au
• Chan, I. and Mostyn, G. (2009). Climatic factors for AS2870 for New South Wales. Australian Geomechanics, Vol 44 No 2. geomechanics.org.au
• Hargreaves, B. (2005). 20 years of AS 2870 in South East Queensland: a personal view. Australian Geomechanics, Vol 40 No 3. geomechanics.org.au
• Li, J., Zou, J., Bayetto, P. and Barker, N. (2016). Shrink-swell index database for Melbourne. Australian Geomechanics, Vol 51 No 3. geomechanics.org.au
• Lopes, D. and Osman, N. Y. (2010). Changes of Thornthwaite's total moisture indices in Victoria from 1948-2007 and the effect on seasonal foundation movements. Australian Geomechanics, Vol 45 No 1. geomechanics.org.au

About this article. Published by LayeredGeo and written from the published research cited in the Sources section above. It is general information about how property and ground conditions are assessed in Australia, not engineering, planning, legal or financial advice, and it is not specific to any property. Check anything that matters against the source dataset or a suitably qualified professional before you rely on it. If you spot something wrong, tell us at hello@layeredgeo.com.au and we will fix it.

← All articles