S.B.M. Beck

Leak detection in pipelines using cepstrum analysis

The detection and location of leaks in pipeline networks is a major problem and the reduction of these leaks has become a major priority for pipeline authorities around the world. Although the reasons for these leaks are well known, some of the current methods for locating and identifying them are either complicated or imprecise; most of them are time consuming. The work described here shows that cepstrum analysis is a viable approach to leak detection and location in pipeline networks. The method uses pressure waves caused by quickly opening and closing a solenoid valve. Due to their simplicity and robustness, transient analyses provide a plausible route towards leak detection. For this work, the time domain signals of these pressure transients were obtained using a single pressure transducer. These pressure signals were first filtered using discrete wavelets to remove the dc offset, and the low and high frequencies. They were then analysed using a cepstrum method which identified the time delay between the initial wave and its reflections. There were some features in the processed results which can be ascribed to features in the pipeline network such as junctions and pipe ends. When holes were drilled in the pipe, new peaks occurred which identified the presence of a leak in the pipeline network. When tested with holes of different sizes, the amplitude of the processed peak was seen to increase as the cube root of the leak diameter. Using this method, it is possible to identify leaks that are difficult to find by other methods as they are small in comparison with the flow through the pipe.

Fluidic Pressure Pulse Transmitting Flowmeter

This paper describes a fluidic flowmeter, consisting of a Coanda Switch and Vortex Amplifier, which has been specially developed for remote flow measurement. The device generates pressure pulses at a frequency proportional to the volume flowrate. The pulses are transmitted via the flowing fluid and are detected downstream. Details are given of the geometry requirements for oil flow measurement in an oil producer, and how the Coanda Switch and Vortex Amplifier can be matched based on the scaling of their characteristics. The predicted value of the flowmeter performance index, N, is found to agree well with experimental results. It is found that this index N has a maximum value when the area ratio of the Vortex Amplifier exit to the Coanda Switch supply nozzle is 0.8. Optimization of the flowmeter device with different control loop designs has also been conducted and has shown good performance over a wide operating range in the laboratory.

Leak Monitoring in Pipeline Networks Using Wavelet Analysis

Leaks and blockages in fluid filled pipeline networks are one of t he major concerns in chemical, gas, oil and water distribution industries. Current monitori g procedures of these networks require the use of multiple sensors distributed throughout the pipes. These include: pressure sensors, flow meters and/or valve sensors. Leak detection is usually accomplished by monitoring the fluid flow rate in and out of a pipe.. Recent work in this area has shown applications of wave reflections identified from the water hammer tests. Thi s paper proposes a wavelet-based procedure for leak severity monitoring. The method detects singula rities in reflected pressure waves. The method is validated through pipe network simulations. Pressure waves are simulated using time domain based transmission line modelling techniques. The st udy shows great potential for this method of leak severity identification which can be applied to many industria l applications. Introduction A number of pipeline network monitoring methods and leak detection techniques have been developed over the last forty years. These developments are prim arily driven by environmental and financial factors. Severe leaks can lead to water shortages or nvironmental catastrophes. Water leaks are associated usually with ageing infrastructure of wa ter distribution systems. Environmental leaks are one of the major concerns in chemical and oil industries. The need for rel atively cheap and simple methods to monitor pipelines is of great interest to all f luid pipeline network owners and users. One of the major disadvantages of current leak detection met hods is the need for the use of multiple sensors. Traditionally, pipeline networks are monitored using var ious types of flow meters, pressure and valve sensors. Flow meters are used to monitor the f low rate at various points of the network. As reported in [1], leak monitoring techniques include acoustic methods, tracer injection and video inspection,. Various methods of signal processing are used to as sis the analysis of sensor data. These include: sensitivity studies [2] and non-linear state observers [3]. Recent developments in the area of leak detection have shown the poss ibilitie of leak detection using wave reflections from water hammer tests [4,5]. The basic idea of this method is to use water hammer to create waves, which propagate through the system. Refl ection waves are captured by a single pressure transducer. Additional reflection points can be rela t d to various features, including potential leaks. More recently, a reliable wavelet-based method for the identification of reflection points in pipeline networks has been proposed [6]. The method has been validated to detect various pipeline networks leaks. The aim of this paper is to extend the study of pipeline network lea k d tection to the identification of the severity of leaks. The paper starts with a brief introducti on to pipeline network modelling. Then a short section describing the wavelet method used for singulari ty detection is then included. Finally, the analysis of simulated results focusing on leak sev erity (simulated by a change in pressure resistance) forms the next step in showing the utility of this metho d.. Key Engineering Materials Online: 2003-07-15 ISSN: 1662-9795, Vols. 245-246, pp 51-58 d i:10.4028/www.scientific.net/KEM.245-246.51

What is a Smart Building

Purpose – Within the building sector a lack of clarity in terminology does not help designers, clients or researchers. Non-domestic buildings have shown rapid increases in the use of advanced technology and control systems with varying drivers, many of which are labelled as intelligent. The term smart has been used interchangeably with intelligent without any clear distinction between the two. If the term Smart Buildings represented a separate, more advanced grouping, it would provide an opportunity to focus the future progress of non-domestic building development. The paper aims to discuss these issues. Design/methodology/approach – Drawing upon academic and industrial literature and experience, this paper reviews the scope of Intelligent Buildings and the current available definitions of Smart Buildings to form a clear definition of both smart and Intelligent Buildings. Findings – These definitions define the border between the intelligent and the (more advanced) Smart Building. The upper bound of the S...

Pipeline system identification through cross-correlation analysis:

Abstract The pipeline systems used to carry liquids and gases for the ventilation of buildings, water distributions networks, and the oil and chemical industries are usually monitored by a multiplicity of pressure, flow, and valve position sensors. By comparing the input signal to a valve with the pressure reading from the network using cross-correlation analysis, the technique described in this paper enables a single sensor to be used for monitoring. Specifically, the offset and gradient change of the cross-correlation function show the time delay between the input wave and the acquired output signal. These reflections arise from junctions, valves, and terminations, which can be located effectively using the cross-correlation technique. Investigations using a T-shaped pipe network have been conducted with a valve inserted in the pipeline to introduce artificial water hammer-type perturbations into the system. Both computational and experimental data are presented and the results are compared with the actual pipe network geometry. It is shown that it is possible to identify the location of various features of the network from the reflections and thus to perform either system characterisation or condition monitoring.

Proportional Feedback Control of an Electro-Rheological Vibration Damper

In this paper, the authors derive a quasi-steady model of an electro-rheological damper under proportional feedback control. Using this model it is shown how, through the choice of two gain terms, the input/output characteristic of the closed-loop system can be linearised. Also, through the choice of a third gain term, the slope of the force/velocity line can be determined. A numerical study compares the performance of openand closed-loop schemes and indicates clearly the working range over which linear behaviour can be obtained.

Measurement of fluid flow rates through cracks

An experimental study has been performed into leak rates through a series of cracked specimens. Semi-elliptic fatigue cracks were grown through aluminium plates and the flow of water was measured as a function of crack length and crack opening. Two relationships were found between volumetric flow rate and crack opening in the region of flow studied. For larger crack openings theoretical results obtained by modelling laminar flow between parallel planes corresponded well to the experimental data. At smaller crack openings, agreement was found with theoretical results obtained by modelling the pressure drop due to bends and changes in area. It was also found that below a threshold value of crack opening no flow occurred.

A remote measuring flow meter for petroleum and other industrial applications

The blown venturi flow meter based on a Coanda switched venturi resistance, has been investigated for use in petroleum and other industrial applications. It produces self-induced oscillations in the fluid at a frequency proportional to the rate of flow and the pressure waves are transmitted via the flowing fluid and detected downstream. Experimental work has been conducted by measuring the attenuation of the waves in air flow along a 28.5 m long, 0.05 m diameter pipe downstream of the flow meter, ensuring that the remote flow metering concept is realized. Predictions of pressure pulse strength, pressure loss and frequency for the flow meter in an oil production string have been made. Potential applications of this meter to nuclear fuel reprocessing and other hostile environments have also been demonstrated. Furthermore, accuracy and limitations of the flow meter for engineering applications are considered.

Instantaneous Phase and Frequency for the Detection of Leaks and Features in a Pipeline System

This work focuses on the progress of a new analysis technique to detect pipeline leaks and features based on analysis of a pressure transient. The measured time domain signals of these transients were obtained using a single pressure transducer. The instantaneous phase and frequency of the signals are analyzed using the Hilbert transform (HT) and the Hilbert—Huang transform (HHT). Both simulated and experimental pressure signals were used to evaluate the performance of these transforms. The analysis of simulated signals allowed features in the pipeline network such as leaks and pipe ends to be ascribed to features in the signals. Analysis of the experimental tests corroborates the simulated test results when the HT and the HHT analysis are used.

Transmission line modelling of simulated drill strings undergoing water-hammer

Drill strings and oil production lines are examples of fluid systems for which time-dependent (dynamic) as well as steady state (static) analysis is increasingly needed. These systems are difficult and expensive to instrument and test experimentally. Developments of fluidic non-moving-part controllers to produce water-hammer pulsations stimulated a need to simulate the fluid dynamics of such drill strings to aid the design work. The method of simulation chosen was transmission line modelling (TLM). It is essentially a time-delay method, borrowing its main concepts and the fundamentals of its computational solution scheme from early work by others in the field of electrical power lines. In its elementary form, a fluid network is treated as a set of pipes (or pipe segments) where waves travel with pure time delay. Connecting the pipes are junctions of various types at which the waves are scattered (transmitted, reflected and/or attenuated). The merits and limitations expected with this methodology in comparison with the method of characteristics (MOC) and other wave-analysis methods are discussed. The first adaptations of TLM were for small perturbation analysis. The presentation here takes such work further forward to model large-scale waves in pipe networks of almost arbitrarily complex topology. The basic theory behind the method is presented and the solution schemes are formulated mathematically with comments on the type of data structure and algorithms needed to undertake computationally such solutions. With the aid of modules described elsewhere, providing comprehensive steady state modelling capability, the software provides a powerful tool for implementing static and dynamic TLM simulations of networks. One of the novel aspects of considerable benefit is the ease of implementation of time-varying junctions capable of representing the overall action of control elements such as the fluidic controllers mentioned earlier. A large experimental laboratory facility with a simple circuit containing the essential hydrodynamics of drill strings was used to gather data on water-hammer pulsations. A controlled solenoid valve with a high-resistance bypass acted as an alternating high and low resistance in the main pipe loop. A simplified version of the circuit was simulated with TLM to compare and discuss the results. The TLM time-domain results took a few seconds of computer processing time and revealed the basic features of the circuit dynamics quantifying water-hammer to a fair and useful accuracy. Such results were encouraging and confirmed the power of this computational method as an aid in the design process.