Soakage testing: modelling the effect of antecedent hydrology

Current design methods for sustainable drainage system infiltration devices, such as soakaways, permeable pavements and infiltration basins, require knowledge of the empirical soil infiltration coefficient as determined using a soakage test procedure. However, the control that seasonal antecedent hydrological conditions, specifically unsaturated zone soil moisture and water table position, have on the infiltration coefficient is uncertain. This paper uses exploratory numerical modelling to investigate this uncertainty. Results show that antecedent water table position has a significant control on the infiltration coefficient, whilst for a properly conducted test the seasonal antecedent soil moisture is less important. To avoid over-estimating the infiltration coefficient, soakage testing should ideally be performed under the most onerous hydrological conditions and three successive drain downs must be used. Model outputs such as those presented can be used to reduce the design infiltration coefficient to account for tests made during periods of lower groundwater level and higher design water table positions. Additionally, the current infiltration coefficient calculation is shown to be biased toward the initial, more rapid, infiltration rate. Therefore, an alternative full depth infiltration coefficient calculation is proposed. Finally a potential detrimental feedback is hypothesised between infiltration device performance and groundwater mounding. Significant design implications are discussed.