Graham Levy, senior technical director Water Resources at Beca, won the 2017 Water New Zealand Stormwater Group Paper of the Year award for these findings which he also presented at the recent Water New Zealand Stormwater 2017 conference in Auckland.
Technical Publication 108 (TP108) was published by Auckland Regional Council (ARC) in 1999 and has been used as the primary flow estimation tool in the Auckland region since then. For the past 10 years there has been consideration by ARC and then Auckland Council (AC) of updating the guideline, but a new version is yet to be adopted.
Similarly, there have been moves within the stormwater industry, supported and promoted by Water New Zealand, to develop a national standard similar to Australian Rainfall and Runoff to establish consistent and reliable flow estimation methods across New Zealand.
In the absence of that guidance, many different flow estimation methods are in use, with variable reliability and suitability.
In the absence of national guidance, TP108 – or the NRCS (Natural Resources Conservation Service of the US Department of Agriculture) method on which it is based – has been adopted and adapted in part or in full for urban runoff estimation in other parts of New Zealand, and is widely used, but also in some respects misused.
This misuse arises in part from a lack of understanding of the basis on which TP108 was adapted to the Auckland Region, and sometimes a lack of robust (or in some cases any) validation of the method to local conditions in new areas.
In his full paper, Graham Levy sets out some underlying principles on which the application of the NRCS method to Auckland were based, and how it was adapted and validated to suit the particular requirements that ARC had defined.
The paper discusses some examples where the method has been used inappropriately, or in new areas, resulting in poor estimation of runoff characteristics. From there, it provides some guidance on factors that should be addressed, particularly in relation to validation, when transferring the method to other parts of New Zealand.
There is also commentary on appropriate contexts in which to use the method, and where other tools might be more appropriate.
TP108 or related methods based on NRCS are being used more widely, but are not necessarily reliable or validated for the particular context where they are being used.
The NRCS method is often misunderstood or misapplied, leading to reduced reliability. If the method is to be used in a new area, there should be a focus on validation to local conditions, including specific consideration of both volume and peak flow rate.
The NRCS method is not suited to all situations – other tools are better in some contexts.
In particular, it is not ideal for high-definition detailed urban models. It is perhaps more complex than necessary for simple site-level runoff calculations where rational formula is simpler and potentially adequate. Neither is it ideal for a continuous simulation context that is needed for understanding lifecycle performance of urban drainage systems and devices.
The ideal would be a nationwide consistency of approach to rainfall runoff estimation that gave the industry and the public greater confidence in reliability.
This approach would include different methods for different design / analysis contexts, but would be based on real flow data and appropriate validation in different contexts across the country.
Consideration should be given to methods that allow for continuous simulation, to better reflect the importance of environmental effects and everyday flows, rather than just for flood estimation.