The challenges of monitoring urban stormwater flows

Peter Stephens - NZ AWT , 22 October, 2010

Over the past 6 years in New Zealand there has been a growing emphasis, particularly in major cities, to develop integrated catchment plans based on the outcomes from potable water, stormwater and wastewater models. Techniques for the monitoring and calibration of water and wastewater flow monitoring networks are well established. The monitoring of urban stormwater flows is still in its infancy with no standardised guidelines or specifications.

The predominant driver for stormwater models is flood hazard assessment. Most existing stormwater models are based on theoretical flows derived from rainfall and established run-off coefficients that are based on land usage. Flood hazard assessment minimises the economic risk associated with the flooding of habitable floor levels. The consequence of getting the flood hazard assessment wrong varies from catastrophic (loss of life) to severe (property damage). Equally important is to not falsely identify a property under flood risk as there are financial impacts to the ratepayer – building restrictions and/or devalued property. Therefore there is a growing emphasis to calibrate stormwater models using stormwater flow data.

Over the past 20 years, urban stormwater monitoring has mainly been undertaken using combined depth / velocity flow monitors. This type of Doppler velocity flow monitoring equipment was primarily designed for wastewater monitoring and is not ideally suited for large stormwater flows. It is also expensive, so monitoring periods are often limited to durations of 6-12 months rather than decades.

Over the past 6 months AWT has installed and operated a 13 station urban stormwater monitoring network in Auckland, New Zealand. Storm flows vary at the sites from 1 – 75 m3. Despite the availability of permanent acoustic technologies we have opted for the low cost and reliability of a stilling well / float system for the continuous monitoring of stage. The reduced equipment costs and relative simplicity has encouraged clients to undertake longer term (up to 5 years) monitoring programmes. GPRS telemetry has been installed at all flow and rainfall stations to provide real-time data via a web-based interface. SMS-alarming of both flow and rainfall stations have enabled field crews to be mobilised for storm gaugings maximising our ability to undertake calibrations during high intensity / short duration storm events.

As we have opted to not continuously monitor velocity, there is a reliance on comprehensive stage-discharge relationships to provide accurate flow data. The rapid stage changes, high flow rates and short event durations of urban stormwater systems provide challenges with developing stage to discharge relationships. It has been found that using a combination of ‘older methodologies’ such as salt dilution gauging and current meter gaugings with the use of acoustic technologies monitors provides the best and safest calibration methodologies.

This paper discusses the challenges that are faced with urban stormwater flow monitoring and the methodologies employed (both successful and unsuccessful) by AWT to obtain the most accurate storm flow data. It highlights that although acoustic technologies are the future of urban stormwater monitoring, technological trade-offs often have to be made to fit client budgets. The use of established stream gauging methodologies and equipment do not necessarily compromise data quality or availability provided that the sites are well calibrated. The use of GPRS telemetry for both SMS alarming and dissemination of the data is highlighted.