Online model- and data-based leakage localization in district heating networks - Impact of random measurement errors

Leakages in District Heating Networks—Model-Based Data Set Quality Assessment and Localization

“…In Pierl’s paper [ 12 ], three different model- and data-driven approaches evaluate artificial measurement data in a DHN for leakage localization. Two methods can only be used within a new steady state after leakage occurrence whereas the third approach evaluates the resulting pressure wave.…”

“…The negative pressure wave is evaluated, which is also the basis for MoFoDatEv for the leakage localization. In [ 10 , 11 , 12 ], the presented evaluation of the pressure wave is divided into two steps: The determination of the PDTPs and the attribution to the EAs. The best suitable algorithm for the PDTP detection is presented in [ 3 ] and the required framework in [ 4 ].…”

“…The same model with the same model assumptions and quality criterions is used for leakage localization. Previous publications have shown that two steps are necessary for the localization [ 4 , 10 , 11 , 12 , 21 ]. The first step is the determination of the PDTP [ 3 ].…”

“…The network investigated here has a total trench length of around 90 km and 8250 pipes. In [ 4 , 12 , 21 ], TLs are placed in a 10 m grid with at least one TL per pipe. Therefore, 24,368 TLs were placed throughout the network.…”

“…For the events 5, 9, 11, and 13, the affected EA is ranked in the first place. Furthermore, the BCP algorithm ranked the affected EA for nine events (5,9,11,12,13,14,19,20,22) in the first three places. However, some events have a worse PC of only 35.7%, which indicates a worse place in the ranking.…”