Pumice soils in the Bay of Plenty are among the most permeable in New Zealand, and designing stormwater systems for them is the mirror image of designing for the Pallic clays of the Rangitikei. At a commercial farm near Otakiri, measured infiltration rates ranged from 800 to 1,720 mm/hr, but the design adopted a conservative 400 mm/hr to account for soil variability, compaction around installed devices, and long-term clogging risk. Understanding how to apply a safety factor to high-permeability soils, without either over-designing or under-designing the soakage system, is a distinct engineering skill.
What Makes Pumice Soils Different
Pumice is a volcanic glass formed during explosive eruptions. The Taupo Volcanic Zone has deposited pumice-rich tephra across a wide area of the central North Island, with particularly deep deposits in the Bay of Plenty, Waikato, and parts of the Rangitikei. The engineering properties of pumice soils are distinctive: low bulk density (typically 600 to 1,100 kg/m3, compared with 1,400 to 1,800 kg/m3 for typical mineral soils), high porosity, and very high saturated hydraulic conductivity.
For stormwater engineering, the relevant property is infiltration rate: how quickly water enters the soil surface and moves downward through the profile. In pumice soils, this rate can be extraordinarily high. At the Otakiri site, field testing using the falling-head permeameter method returned rates of 800 mm/hr at the least permeable test location and 1,720 mm/hr at the most permeable. For context, a typical clay soil in the Manawatu might return 2 to 10 mm/hr. The pumice is two orders of magnitude more permeable.
This creates a design problem that is counterintuitive. With soils this permeable, the stormwater system should be simple: direct runoff into a soakage device and let the ground absorb it. In practice, it is not that straightforward.
Why You Cannot Design to the Measured Rate
Field infiltration tests measure the soil's performance at one point, on one day, under controlled conditions. The design must account for conditions that differ from the test in several important ways:
- Spatial variability: Pumice deposits are not uniform. A test pit 5 metres away from the first may encounter a layer of fine volcanic ash or a palaeosol (buried soil horizon) that reduces the infiltration rate by half or more. At the Otakiri site, the measured range of 800 to 1,720 mm/hr across the site demonstrates this variability directly.
- Construction disturbance: Excavating a soakage pit or installing a soakage chamber compacts the surrounding soil. Tracked machinery working on pumice can reduce the near-surface permeability by 30 to 50% within the zone of compaction. The soakage device is installed in disturbed ground, not in the undisturbed soil that was tested.
- Long-term clogging: Fine sediment, organic matter, and biological growth progressively reduce the effective infiltration rate of any soakage device over its design life. In pumice soils, the large pore spaces are particularly vulnerable to blockage by fine material washed in from impervious surfaces. A pre-treatment device (such as a silt trap or filter strip) mitigates this, but does not eliminate it.
- Groundwater mounding: At very high infiltration rates, the rate at which water enters the ground can exceed the rate at which it disperses laterally through the aquifer. This creates a temporary mound of groundwater beneath the soakage device, which reduces the effective head driving infiltration and slows the apparent soakage rate.
Applying the Safety Factor
At the Otakiri site, the design infiltration rate was set at 400 mm/hr: approximately half the lowest measured rate and less than a quarter of the highest. This is a safety factor of 2.0 to 4.3 depending on which test location is used as the reference.
The choice of 400 mm/hr was not arbitrary. It was derived by applying three independent reduction factors to the measured minimum of 800 mm/hr:
- Construction disturbance factor: 0.7 (a 30% reduction for compaction effects around the installed device)
- Long-term clogging factor: 0.8 (a 20% reduction for progressive loss of permeability over the device's design life, assuming pre-treatment is provided)
- Variability factor: 0.9 (a 10% reduction to account for the possibility that the soakage device is installed at a location slightly less permeable than the test pit)
Multiplied together: 800 x 0.7 x 0.8 x 0.9 = 403 mm/hr, rounded to 400 mm/hr for design. Each factor is individually justifiable, and each can be reviewed and adjusted if additional site information becomes available.
This approach is more defensible than simply halving the measured rate or applying a blanket factor of safety. A reviewer can ask "why 0.7 for construction disturbance?" and receive a specific answer grounded in published literature on compaction effects in volcanic soils.
The Contrast with Low-Permeability Sites
The engineering skill required for pumice soils is the inverse of what is needed on the Pallic clays of the Rangitikei, where SAE has designed multiple subdivisions. On clay sites, the challenge is that soakage does not work at all: infiltration rates of 2 to 10 mm/hr mean that soakage devices would need to be impractically large, and the design defaults to attenuation and controlled discharge to a receiving environment.
On pumice, the challenge is that soakage works too well in the test, and the designer must resist the temptation to use the raw measured rate. Over-designing (using 100 mm/hr when 400 mm/hr is defensible) wastes money on oversized devices that the soil does not need. Under-designing (using 1,000 mm/hr because "that's what the test showed") creates a system that may fail when any of the reduction factors listed above come into play.
The correct approach is in the middle: a design rate that is conservatively derived from measured data, with each reduction factor documented and traceable.
Device Selection for High-Permeability Soils
At the Otakiri site, the stormwater system used soakage pits lined with geotextile and filled with clean aggregate. The geotextile serves as a filter, preventing fine sediment from migrating into the surrounding pumice and progressively clogging the soil pores. The aggregate provides void space for temporary storage during peak rainfall, and the high infiltration rate of the surrounding soil ensures rapid drainage between events.
The device sizing was straightforward once the design infiltration rate was established. At 400 mm/hr, even a modestly sized soakage pit can handle the runoff from a large impervious area. The facility included approximately 3,500 m2 of new roofing and hardstand, and the total soakage pit volume required was significantly less than it would have been on a site with typical mineral soils.
Pre-treatment was specified upstream of the soakage pits: silt traps on hardstand runoff and leaf screens on roof downpipes. These measures protect the soakage device from the fine sediment loading that is the primary long-term failure mechanism for soakage systems in any soil type.
High measured infiltration rates do not mean the stormwater design is simple. At the Otakiri site, the design adopted 400 mm/hr from measured rates of 800 to 1,720 mm/hr, using three independently justified reduction factors. The result is a system that is conservatively designed without being over-engineered.