There are few common construction systems more problematic than waffle rafts. Scott Chen, a structural engineer,  has raised his unreserved structural concerns in his Linkedin post and discussion (unfortunately the post, and the attached knowledge-base from the spirited discussion, has been removed by LinkedIn). The structural issue arises from the system being “completely on-ground rather than in-ground” (AS2870-2011 Residential Slabs and Footings, Section C3 ‘Standard designs’). The lack of pier engagement becomes a particular problem when building on reactive clay soils.

There is also another major drawback with waffle rafts, these slabs are fiendishly difficult to damp-proof. Once again, this is not an unknown limitation, the Standard acknowledges that  “there is a greater propensity for ingress of moisture under the slab.” Nevertheless, I fear that the industry, and particularly building occupants sensitive to mould, are not cognizant to the likelihood of this becoming an unmitigated disaster.

First off I should note that there are differences in definitions of the term “waffle raft” between  AS2870-2011 and the National Construction Code, NCC 2019. Legislatively speaking, the NCC overrules in such differences, but I should call out that the NCC definition is technically incorrect.

Waffle raft means a stiffened raft with closely spaced ribs constructed on the ground and with slab panels supported between ribs. (NCC 2019 Schedule 3 ‘Definitions’)
Waffle Raft. A stiffened raft with closely spaced ribs constructed on the ground and with slab panels suspended between ribs. (AS28790-2011, 1.8.63 ‘Definitions’)

What is the difference between ‘support’ and ‘suspend’? To illustrate, we would say that above-ground electric cables are suspended between telegraph posts. It is incorrect to say they are supported between posts. I would prefer not to have to be such a pedant, but precision is important, and this difference in definition becomes important at a later discussion in this article. For present purposes, I will be using the AS2870 definition of a waffle raft.

Conventional stiffened rafts are constructed first by excavating the form for the beams, and often the slab as well, in-ground. Waffle rafts are perceived as a more cost effective way of casting an equivalent slab by constructing on-ground. Void-formers are arranged in a grid, with reinforcement placed between those void-formers to form the ribs of the waffle.

Left: EPS Waffle pod manufacture (source: Pod Slab Sales).
Right: Stackable plastic pods (source: Neumann Steel)

The NCC 2019 requires that in most states a vapour barrier of medium impact resistance is to be laid under the slab. In NSW, a damp-proofing membrane (DPM) of high impact resistance is specified. In either case, these membranes are supplied at typical widths of 4m and will need to be overlapped at the seams/joints. The NCC specifies the overlap be a minimum of 200mm, with no requirement for taping or other form of adhesion at these seams.

NCC 2019 Volume Two

For the most part, the absence of sealing the overlapped joints does not appear to have been a problem for conventional flat slabs and stiffened rafts. After all, the joint is continuously sandwiched between the ground and the weight of the concrete slab. Observe in the AS2870 illustration how the DPM is immediately against the soffit (i.e. underside) or the slab.

AS2870-2011. Residential slabs and footings. Use of DPM for slab protection.

The same assumption cannot be made with waffle rafts. Observe how the DPM is under the void and not against the soffit of the slab.Critically, if the seam appears under a void (instead of under a rib) there is nothing to hold the overlap against a high water table event where water is pushed up against the membrane by hydrostatic pressure. 

Water ingress into a waffle raft.

Presented graphically:

  • A: The seams, when located under a void will have no pressure applied so as to keep the DPM sealed.
  • B: When the slab is cast and the weight of the building is applied onto the foundation (i.e. soil), there is differential settlement due to the loads being transmitted through the ribs, and the absence of it in the voids.
  • C: If the ground becomes saturated during a heavy downpour, the water table can applying hydrostatic pressure, emerging above the DPM into the void.
  • D: When the water table recedes, water is trapped in puddles above the DPM. The slab is sitting in these puddles, creating rising damp in the waffle, and high humidity in the subfloor. The DPM thus becomes a one way system for water to enter but not completely leave.

The differential settlement in B is a concept that I should elaborate. This is where I had to be pedantic about definitions. Void formers, or pods, are either made of EPS (expanded polystyrene) or recycled plastic. These pods are simply there to create voids, and do not ‘support’ the concrete floor. People who applied point loads on waffle raft slabs have found out the hard way that they could punch through the slab when jacking a car up on the hollow parts of the garage slab. Thus the weight of the concrete, the weight of the structure (dead load) and fluctuating live loads from occupants and wind, are all completely borne by the ribs and edge beams. The ground under these rib structures are typically not compacted and will settle over time, whereas those unloaded areas under the voids will not settle. The underlying  membrane thus becomes stretched. It is unclear if taping these joints will at all ensure membrane continuity under such stretching, or even if the membranes maintain the required 200 micron thickness when stretched, or if they rip under such substantive loads.

Since the DPM cannot be assured to have continuity, then the ground around and under the building should never be allowed to get saturated, so as never to apply hydrostatic pressure on the membrane. This can be achieved by diligent and fail-safe over-engineering of site and subsoil drainage. This gets increasingly challenging as sites become tighter and the amount of permeable ground surface is restricted. But when it comes to this membrane acting as a vapour barrier, it is near impossible to ensure that the ground will never apply vapour pressure on the membrane so as to create high humidity in the voids.

I have been thinking of a way to remedy this problem endemic to the industry today. I still do not have a good solution. So my best recommendation, one that would go without saying, is that this problem should have been avoided in the first place by using a different footing system. 

Who would have thought that those 200 microns of membrane can be so important, and so hard to rectify after construction?