Submitted to: International Conference on Water Resources Engineering Proceedings
Publication Type: Proceedings
Publication Acceptance Date: August 7, 1998
Publication Date: N/A
Interpretive Summary: Water is becoming a scarce resource, and agricultural water users are under pressure to use it more judiciously. For many large irrigation projects, the physical infrastructure that delivers water to users influences their ability to manage the water supplied to them. Most large water delivery systems convey and distribute water with canals rather than pipelines. Infrastructure improvements (e.g., conversion to pressurized pipelines) ar typically very expensive relative to changes in operations. Operations can be improved by providing canal operators with better tools for determining control actions. One such tool is computerized automatic control of canal gates. This technology, however, is not routinely available to irrigation district personnel and consultants. In 1997, the Salt River Project (SRP) established a pilot project on canal automation. Under this project, the U. S. Water Conservation Lab is assisting SRP with the testing and implementation of canal automation on a small portion of their system. This paper describes progress on integrating this automation system into their existing operations. If successful, the automation system will be expanded to SRP's entire network. The results of this study should be of interest to the U. S. Bureau Reclamation, department of water resources in western states, consulting engineers, and irrigation district personnel.
Technical Abstract: Canal automation has found its way into practice under two situations: local automatic control of a single structure (e.g., flow rate control at the head of a lateral) and control of volume for large schemes based on pre-scheduled or predicted demands (e.g., Dynamic Regulation). To date, feedback control of a series of canal pools has proven to be very difficult, with a few marginal successes (.e.g., ELFLO on the Corning Cana in CA) and several dismal failures. In our view, there are three reasons why canal automation has not found its way into practice. First, the feedback control algorithms in use were not sophisticated enough to handle the interaction between pools. This technical problem has not been solved. Second, the canal's hydraulic properties limit the ability of feedback control to handle typical demand changes. This constraint becomes more severe as flow changes become larger. Anticipation and prescheduling of large demand changes is usually necessary. New gate-stroking procedures have recently been developed than can be efficiently implemented. Third, the implementation of the feedback control algorithms have been done in isolation from the rest of the system. In the current scheme, feedback, feedforward, and manual control of a canal can occur simultaneously, thereby utilizing the advantages each has to offer. This system is currently being installed for testing on Salt River Project's Arizona Canal.