Investigation of Active Identification Methods for Thermohydraulic Testing of Heat Networks

Resumo

Considerable uncertainty of the information about actual characteristics and parameters is the key factor restricting the effective use of mathematical modeling and computer simulation methods in solving problems of optimal retrofitting, commissioning, development of operating regimes, and dispatch control of heat networks. Application of the existing industrial procedures in testing a heat network for hydraulic and heat losses does not make the issue less urgent, because of the fact that the test conditions have not been properly standardized as yet and no guarantee is provided for the completeness and accuracy of the obtained results. The employment of the available methods of parametric identification during passive monitoring of heat networks during their normal operation does not ensure that the required solution will be obtained due to insufficient number of measurement points and a narrow range where the operating conditions can be varied. The paper presents a unique formalization of the problems of testing heat networks for hydraulic and heat losses as problems of active identification assuming optimal planning and processing of the test results with the use of a mathematical model of steady-state thermohydraulic regimes. A description of the model, a substantiation of selection of test optimality criteria, mathematical formulations of problems for optimal planning of test conditions and optimal location of instruments, and test result processing procedures are presented. The proposed procedure is in a step-wise test strategy giving maximum information with a minimal risk that too many tests will be performed. It is applicable for testing various types of heat networks having any structure or configuration. The potential efficiency of simultaneous performance of thermal and hydraulic testing of a heat network is demonstrated for the first time theoretically and by an example. It manifests itself in minimizing the total number of experiments required for reaching the prespecified accuracy in determining the actual characteristics of heat networks and of the model predictions.