Abstract

Melbourne Water is a Victorian government owned company that provides water supply and sewerage services to Melbourne's retail water companies, and manages rivers and creeks, floodplains and the regional drainage system. The water supply system managed by Melbourne Water is a complex interconnected system of 10 storage reservoirs, over 40 service reservoirs, 160,000 hectares of catchments and hundreds of kilometres of tunnels and pipelines. The complexity of the system and the needs of stakeholders mean the way the system operates has to be based on many considerations. These considerations include: the volumes to be sourced from various sources that have differential costs, the timing of volumes sourced, the distribution of storage within the water supply system, minimising the risk of localised or system wide spills due to increased supply from various sources, minimising the risk of low storage volumes due to inadequate intakes volumes, water security, water quality, scheduled outages of system assets, operating and capital costs, and environmental factors such as healthy water ways and lower greenhouse gas emissions. In planning for the system operation, Melbourne Water prepares an annual operating plan (AOP) that provides a forecast of anticipated system operation for the water supply system. The AOP outlines the system operations for the 12 months ahead, and identifies for each month the volumes of water to be sourced, stored, moved from one location to another, or released to waterways. The preparation of this forecast is a complex task, as the decisions on system operations need to be made under many considerations as outlined earlier. In the past, the preparation of AOP was a manual process undertaken by experienced operators assisted by decision support tools based around the calculation of water balance across the system. However, these tools did not have the capability to identify optimal operations. Instead, the tools were manually updated by experienced operators as documented in system operation guidelines, in consultation with Melbourne's retail water companies. The above approach was time consuming, often requiring up to two weeks to produce an AOP for one operating scenario. In 2009, Melbourne Water developed in-house an optimisation modelling tool, OPTIMISIR, to assist in identifying the optimal annual operating plan. OPTIMISIR provides the flexibility for users to specify various streamflow and demand scenarios, system storage conditions, operating requirements and the aims of the optimisation in the form of an objective function. The choice two optimisation methods; Linear Programming or Quadratic Programming is also available. The output of OPTIMISIR comprises the operating decisions for each of the 12 months including: volumes to be taken from various water sources, volumes to be released to waterways, volumes to be stored in each reservoir and volumes to be transferred across the system. OPTIMISIR has two modes of use; standalone-use and the linked-use with Melbourne Water's REALM water supply system simulation model. This paper covers the standalone-use mode. With OPTIMISIR, the optimisation of 12-month operations can be completed in seconds. This enables the operators to generate at first a 'greenfield solution' that is not biased towards past operations and subsequently assess the impacts of various operating considerations by generating 'constrained solutions'. The paper presents the OPTIMISIR modelling approach and an example application. The paper particularly highlights the decision-support rather than decision-making nature of OPTIMISIR, in which the experience of system operators and the power of an easy-to-use modelling tool are combined in identifying a detailed operating plan that meets various operating considerations and the needs of stakeholders. The paper demonstrates the suitability of OPTIMISIR, as part of a suite of optimisation and simulation modeling tools, for the operation planning of Melbourne's water supply system.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call