Improving leak detectability in long liquids pipelines

In the pressure-flow deviation method of leak detection, real-time data of pressure, temperature and flow rate are acquired by the supervisory control and data acquisition (SCADA) system at both ends of the pipeline segment under consideration. A real-time model (RTM) simulates the current pipeline flow conditions, taking one boundary condition at each end of the pipeline segment from the SCADA system. Since the real-time model does not account for leaks, discrepancies between the RTM quantities and the SCADA corresponding quantities may be attributed to leaks. There are three possible combinations of boundary conditions used by the real-time model to perform computations.; Three computer codes were written to simulate the different combinations of boundary conditions mentioned above. The resulting models are the QH-model, the HQ-model and the HH-model. A fourth computer code was written so that it uses a mixture of the discrepancies of the QH-model and the HQ-model and makes use of the advantages of both models. This is the QHQ-model. The models were tested under different flow conditions and pipeline parameters. The QH-model resulted in maximum sensitivity (i.e., maximum capability of detecting small leaks) to leaks which take place in the vicinity of the upstream end of the pipeline. The HQ-model resulted in maximum sensitivity to leaks which take place toward the downstream end of the pipeline. The HH-model and the QHQ-model resulted in maximum sensitivity to leaks which take place toward the middle of the pipeline. The leak detectability for all the models increased with increasing pipe flow rate and head loss. To further enhance leak detectability in the discrepancy methods, a new methodology for automated real-time threshold values was established. This new technique yielded threshold values that allowed screening out small leak-caused discrepancies from noise-caused discrepancies, leading to excellent leak detectabilities.; To improve leak detectability for modern leak detection systems, a new approach for leak detection was proposed. The approach is based upon the fact that if a leak is present in the system and the same leak size is imposed upon the real-time model at the exact leak location, the discrepancies are minimal. This technique was more capable of detecting small leaks than the discrepancy methods. A comparison between this method and the discrepancy methods was made under different flow conditions and pipeline parameters.