Automatic Calibration Module for an Urban Drainage System Model

Ivar Annus,Anatoli Vassiljev,Katrin Kaur,Nils Kändler

doi:10.3390/w13101419

Abstract

The purpose of the study was to present an automated module for the calibration of urban drainage system models. A prepared tool based on the Open Water Analytics toolkit included 12 additional calibration parameters as compared to the existing similar solutions. The module included a gradient optimization method that allowed adjustment of up to five parameters simultaneously, and a trial-and-error method that provided the possibility of testing one or two parameters. The user interface was built in MS Excel to simplify use of the developed tool. The user can select preferable parameters for calibration, choose the optimization method, and determine the limits for the calculated values. The performance and functionality of the automatic calibration module was tested in two scenarios using the drainage model of a 10 ha heavily developed area in Tallinn, Estonia. The calibration results revealed that the maximum deviation between the modelled and measured flow rates was less than 5% for both cases. This is a reasonably good fit for drainage models, which typically encounter numerous uncertainties. Therefore, it was concluded that the module can be successfully used for calibrating hydraulic models created in SWMM5.

Highlights

Climate change is having a considerable impact on urban areas [1], bringing more extreme rainfall events [2,3]
“Smart” drainage systems, i.e., real time controlled (RTC) facilities that emerged with the development of information and communication technology (ICT), and integration of low impact development (LID) solutions, have the potential to make the needed changes [6]
Del Giudice and Padulano [20] implemented the GANetXL, an optimization add-in for MS Excel, that was originally developed for water supply networks [22] for the automatic calibration of the hydraulic model

Summary

Introduction

Climate change is having a considerable impact on urban areas [1], bringing more extreme rainfall events [2,3]. The intensities of the extreme rainfalls have increased more than 30%, as measured by rain gauges in Tallinn during the last five years, and are exceeding design thresholds of design standards This means that existing urban drainage systems (UDS) might be incapable of handling the increasing volumes, resulting in surcharge and consecutive pluvial floods. RTC of a drainage system depends directly on the existence of calibrated and validated hydraulic models [6]. Site-specific data concerning the structure and geometric properties of a UDS section is often incomplete, thereby hindering the specification of suitable parameters [10] Most of these cannot be measured directly and providing estimates is a challenge. Del Giudice and Padulano [20] implemented the GANetXL, an optimization add-in for MS Excel, that was originally developed for water supply networks [22] for the automatic calibration of the hydraulic model.

Materials and Methods

The initial and final soil infiltration capacities

Objective function OF

Conclusions

Findings

Methods