Overview Objectives

The project’s main objective was to appraise the situation in Zhanghe and modify the IMSOP and ASSET MANAGER models to include features peculiar to Chinese irrigation schemes.

Project Background and Objectives

Irrigation plays a pivotal role in China’s plans to meet future food demand. But the volume of water available for irrigation is under threat, largely from the increasing thirst of the country’s urban and industrial sectors. In the Zhanghe river basin in China’s Hubei province there is strong interest in widespread introduction of water-saving irrigation (WSI) techniques, which can increase food production using less water. But effective introduction of WSI depends on adequate operation of the water supply system.

Similar problems in Vietnam led Australian scientists from Melbourne University, under the leadership of Professor Hector Malano, to participate in an ACIAR project to study system-wide water management in that country’s irrigation schemes. They worked with Vietnamese agencies to adapt a computer model, IMSOP (Irrigation Main System Operation), to analyse and improve operations and develop the infrastructure and institutional arrangements for pricing irrigation supply services. The team also modified and adapted the computer model ASSET MANAGER to speed up collection, retrieval and analysis of asset data.

The project’s success caught the attention of the Zhanghe irrigation authorities, who approached the Australians on the team to see if they could undertake a similar study in China. Together they developed this small ACIAR project, designed to improve main system water management in China through a demonstration study in the Zhanghe Irrigation Scheme (ZIS).

Progress Reports (Year 1, 2, 3 etc)

First Year 1 (01/07/2002-30/06/2003)
A diagnostic study and description of the system was conducted which revealed details of the operation practices in the Fourth Main Canal, the focus of this project. The canal is operated under an arranged demand schedule with farmers requesting water deliveries from the agency as required. The initial analysis shows that farmers are prone to delaying water orders on the expectation that rain will reduce their water cost which is charged on a volumetric basis. A congestion of orders occurs when farmers submit their requests within a short period of time to avoid crop stress which the system is unable to deliver within its existing capacity. Generally, the Fourth Main Canal receives water on average 4 times a year each time for a period of between 5 and 15 days. Farmers must place water orders with 3-day notice to the canal station which aggregate the farmers’ orders. Canal station demands are then aggregated by the Main Canal Office to determine the total canal inflow required.
The computer model IMSOP is being modified and adapted to handle Zhanghe’s mode of operation. This task involves the collection and upload of input data and the addition of a utility for prediction and sequencing of farmer’s orders. Model data has been collected and processed from existing databases available from the Tuanlin weather station. Data quality checks resulted in the identification and correction of many inconsistencies. A detailed survey on the longitudinal and cross sections for the East Branch canal was carried out in late 2002.
The English version of IMSOP has been installed on ZIS’s computers for training and familiarisation with the model by the agency staff. A Chinese version of the model being prepared by Wuhan University scientists will soon replace the English version. Upon completion of the IMSOP data files, and monitoring of the system during the 2003 irrigation season, a detailed supply-demand analysis of the system will be carried out to complete the diagnostic analysis of the system’s operation.
A survey of farmers was also conducted to obtain their perspective on the operational aspects of the ZIS. The main issues identified by farmers revealed conflicts within WUAs forcing farmers to bypass the WUA and deal directly with the canal stations; existing cross regulators at the upstream end of the canal are inadequate to control water flows and irrigation decisions based on empirical experience rather than on technical principles.
An initial trial of Artificial Neural Network’s (ANN) for predicting water orders by farmers was undertaken during this year in which several issues associated with the use of ANN were examined including the number of nodes and variables in the input layer, number of nodes in the hidden layer and learning cycles. Preliminary trials included a two-input- node-one-output-node network and a one-input node-one-output node network. These networks were shown to describe the general trend of farmers’ order although the ability to predict individual order events was less satisfactory. Further trials will be conducted using different variables and network configuration to improve the accuracy of prediction.
An existing GIS of the Zhanghe system was evaluated and found to be non-geo-referenced. A new GIS is being developed by re-digitising after registering the ZIS maps against GPS based ground observations. This GIS will form the basic platform for the asset management software and a financial model of asset condition and replacement needs will be constructed.
Melbourne University staff has conducted training for Wuhan University staff and Zhanghe staff on the use of GPS for asset data collection, modelling with IMSOP and data collection and processing for IMSOP during this phase of the project. In addition, Wuhan University staff provided field training for Zhanghe staff on flow monitoring and structure rating.

See Final Report

Project Outcomes

The project team found that farmers tended to delay their water orders on the expectation that rain would reduce their water bill (which is charged on a volumetric basis). This led to a congestion of orders when farmers all realised their crops were in danger of water stress and therefore submitted their requests within a short period of time. At that point the system was unable to deliver sufficient water for all.
The team modified and adapted the IMSOP model to account for this mode of operation, resulting in the addition of a utility for prediction and sequencing farmer’s orders. Other IMSOP modifications came from data collected and processed from the Tuanlin weather station’s databases. Through quality checks the team identified and corrected many inconsistencies.
It emerged that changes in the water pricing policy in recent years had led to reduced water demand from farmers and a shortfall in revenues from water fees in relation to cost of water supply. The ASSET MANAGER analysis allowed the irrigation company to calculate actual operational cost of the Fourth Main Canal and develop a sustainable water fee policy.
Wuhan University scientists translated the modified versions of IMSOP and ASSET MANAGER into Chinese and they are now installed on ZIS’s computers. China’s National Centre for Irrigation and Drainage intends to promote the work at ZIS to other irrigation areas in China. Such guidelines will be vital as the Chinese Government tackles the massive effort to rehabilitate and modernise ailing structures.
By facilitating more widespread adoption of WSI practices the project will also help to address problems of increasing water shortage and competition that are prevalent in vast areas of China, especially north of the Yangtze River. In several regions, the lack of water may limit future economic development.

Project ID
Project Country
Inactive project countries
Commissioned Organisation
University of Melbourne, Australia
Project Leader
Associate Professor Hector Malano
03 8344 6645
03 8344 6215
Collaborating Institutions
Wuhan University, China
Project Budget
Start Date
Finish Date
Extension Start Date
Extension Finish Date
ACIAR Research Program Manager
Dr John Skerritt