Water Journal : Water Journal April 2012
smart systems water APRIL 2012 87 state of the water (or waste) as well as the infrastructure. For example, if a home water collection system such as a rainwater tank is full, and there is a rain event forecast, we will need an intelligent system which, upon receiving this spatial and temporal information, can implement actions which discharge some (or all) of the existing tank water for replacement by the expected rainwater; this optimises water quality, reduces flood and prevents the discharge of 'better' quality water. Similarly, if a home waste treatment or septic system is malfunctioning and requires maintenance or pumping out, the relevant water utility and home occupant need the data or temporal and spatial information in order to implement processes and actions which initiate repair of the system or pumping out (if required). Such a smart system is essential in the case of a large number of onsite systems to keep them operational by ongoing notification to the operator remotely. This saves resources and prevents possible health risks from overflow by either 'fixing' the problem before it gets out of hand or taking no action. The case is analogous for sewer mining, where processes and actions can be implemented in a timely manner by initiating a process only when the levels of particular components are appropriate for extraction to be economical. An example of where real-time monitoring of both infrastructure and water quality would have made a difference is the Bellevue Hill landslide (Figure 4). In June 2009, a major cast-iron water main burst in Bellevue Hill, in the eastern suburbs of Sydney. The pipe had been leaking underground for at least 60 hours before rupturing. Indeed, maintenance crews falsely concluded that water coming out of the ground was stormwater; in addition, it took around 12 hours to find the nearest valves in order to isolate the catastrophe from the system. As a result of the rupture, there was a significant landslide that swallowed two cars and a power pole. The land slippage created a 25-metre-deep abyss, which sent tonnes of sand and soil into Cooper Park. The pictorial evidence of the event is shown in Figure 4. In addition, the landslide caused the rupture of a gas main, which sparked a lockdown of the surrounding neighbourhood and closure of Bellevue Hill Public School the following day. How could a real-time monitoring system/intelligent network have assisted (or prevented) this occurrence? • To the extent that intelligent technologies prevent or reduce the frequency of these disruptive or harmful events, or aid in the efficient and effective management of such events, the community benefits from the avoidance of associated direct and externality costs; • Direct monitoring of the pipe by, say, an optical fibre-based array sensor system could have picked up either the leak (change in temperature around the pipe or change in background sound) or stress of the pipe due to loss of support prior to catastrophic failure; or the subtle change in pressure, due to the leakage over the 60 hours prior to the catastrophe; • More frequent placement of more sensitive pressure and flow monitoring equipment would have noted the changing trend of pumping needs and/or the subtle changes in flow and pressure around the affected zone; • Electronic tagging and monitoring of appurtenances would have allowed the immediate location of the nearest valve to be known and turned off, or if modern e-valve systems were installed, it could have automatically closed upon the sudden loss in pressure and increase in flow. What Information is Needed A common and often expensive phenomenon in large water distribution mains is burst pipes. This is often considered unavoidable because of the age and location (i.e. buried) of the infrastructure. The costs of disruption in service delivery have never been fully quantified because of the variation in flow-on effects of such an event, due to the dependence of the costs calculation on the location of the pipe or burst and the population type (i.e. business or household). Rarely does such an event occur in the absence of some induction or initiation process such as a slow leak, cracking or unusual stresses on the pipe, corrosion of the pipe, or change in internal water pressure. If we had spatial and temporal information on the condition of the pipe, reflecting the possible induction processes, potential bursts could be proactively managed and thus prevented. Stormwater and sewerage pipe infrastructure can be similarly monitored with respect to bursts and blockages. In addition, environmental and health risk-driven regulatory needs relating to inputs to and discharges from the system can be addressed, given the availability of spatial and temporal information on critical water quality parameters. In the case of the drinking water system, knowledge of the quantities of water consumed and lost within the system at any given time and location will allow the quantification and identification of the location of losses and, thus, this phenomenon can be managed. This will ensure service delivery requirements are met while at the same time preventing the loss of an increasingly critical resource, as well as maintaining the asset (as the loss is likely to have arisen from inadequately managed and damaged or faulty infrastructure, which can then be repaired or replaced). With knowledge of the quantities of water at any given time and at any given location within the system, demand may be better managed (i.e. optimised) and predicted in relatively simple ways in real time. Similarly the spatial and temporal information on the status of valves and fire hydrants in this system will allow more precise control of flows and proactive management of events such as failures or interruptions due to fire fighting requirements. Figure 3. Sensored modern/future water network. (a) (b) Figure 4. Photographs of the Bellevue Hill water main burst in 2009.
Water Journal May 2012
Water Journal December 2011