Bio & CV
Theme 1 through 3 are supervised by Akira Koizumi (Prof. Emeritus, Tokyo Metro. Univ.) and Yasuhiro Arai(Assoc. Prof.Tokyo Metro. Univ.)
The main objective of water pipelines management is to minimize Life Cycle Cost (LCC), the sum of long-term costs occurring in the Water Distribution Systems (WDS). Pipelines incur a series of costs such as installation, losses due to pipe breaks, and removal. In WDS, sufficient pressure and velocity are hydraulic needs to be met over the planning period. Furthermore, pump operational costs depend heavily on pipe diameters; thus these must be included in the LCC model. Still, some research topics related to LCC remain and need to be further explained: e.g., present value, carbon dioxide emissions, and disaster risk such as the earthquakes.
On the other hand, the ideal combination of pipe replacement time and diameter at the time of replacement is virtually incalculable. Hundreds of thousands of calculations must be done even in the small water distribution systems. As WDS becomes larger to any extent, the calculation will take an almost infinite time to perform. Therefore, efficient computer algorithms such as Genetic Algorithms are needed to optimize water distribution systems in a reasonable amount of time.
Our research group has published some ideas, models, and literature reviews on this research topic.
We tend to associate water supply systems with drinking, sanitary or industrial purposes. However, water supply systems also play a crucial role in fire suppression using hydrants, cisterns, surface water, or all of those. Water companies install the hydrants and cisterns while the fire department pays costs needed for the installation because hydrants and cisterns will be used for fire suppression.
In firefighting situations, unsusally high amount of water flows to a single point (i.e., fire ignition site). Water distribution systems tend to be oversized because of this unexpected and instantaneous fire flow needs in such situations. Likewise, the fire department should finance these oversized pipe installation, but no theoretical basis to easily quantify these additional costs in water distribution systems design has been demonstrated yet. Our research group is developing a series of methods to quantify these additional costs, using the Genetic Algorithms.
We want to clarify the following topics regarding firefighting capacity on water distribution systems design:
(1) correcting the current fire flow requirements based on population or building sizes
(2) dependence of system size and topology on additional costs to be borne by the firefighting department
(3) simplified but realistic hydraulic simulation considering fire flow
We eventually aim to contribute to the modification of the current design criteria.
Theme 1 and 2 above use Genetic Algorithms for water distribution system plans. Such algorithms have often been applied worldwide in combinatorial water distribution systems design for the last three decades. If these optimization algorithms can more reliably find a global optimum solution in a short time, application to larger systems can be extended.
We are trying to apply optimization algorithms being used in other fields, modifying them to WDS contexts, and compare these to well-known algorithms e.g., Simple GA etc. In addition, we tested parallelization to expand applicability to larger water distribution systems. This research topic was often discovered by chance while seeking the solutions for larger objectives such as theme 1 and 2.
