Foreign countries started in the 1960s in pipe network modeling. In the 1980s, with the development of computers and related technologies, the application of telemetry and remote transmission equipment entered the practical stage, and many domestic water supply enterprises have realized water supply network modeling. Modeling of the water supply network system is to establish a mathematical model for simulating the dynamic real-time operation of the network system. The establishment of a water supply pipe network model can keep abreast of the operation status of the pipe network, analyze the bottleneck of the pipe network operation, and at the same time play a guiding role in the optimization, reconstruction and expansion of the pipe network, accident handling analysis, etc., and can be further applied to the leakage control of the pipe network, water pump Optimal scheduling and water quality analysis can play an important role in the actual work of water supply companies.
1.1 Basic theory of modeling
The water supply pipe network system is a network system with complex topology, large scale, strong randomness of water use change, and multi-objective operation control. In the past, the management of the buried water supply pipe network was mostly empirical, and experiments and a large number of tests could not be directly carried out. It was very difficult to realize scientific and modern management. Pipe network modeling is by far the most effective method for simulating the dynamic conditions of a water distribution network system. It can provide valuable information and help realize the scientific and modern management of the pipeline network.
Modeling of water supply pipe network system is the process of establishing a mathematical model for simulating the pipe network system. The simulation content mainly has three aspects:
(1) Graphical simulation: On the basis of feasible simplification, the complex topological structure of the water supply pipe network system is input into the computer to realize the simulated pipe network graph, including water source pipe sections, pipe lengths, pipe diameters, nodes, valves, fire hydrants and other accessories , establish a graphic database of the pipe network from these static data of the pipe network.
(2) State simulation: including the time-varying flow of pipe network nodes; the amount of pipeline leakage; the influence of the water level of high-level pools over time; and the opening degree of valves. The purpose of the state simulation is to establish the static and dynamic data of the pipe network and conduct the hydraulic analysis of the pipe network by solving the state equation of the pipe network.
(3) Parameter simulation: Calculate and simulate parameters that do not change with time, such as the resistance coefficient C value of the pipeline, the influence of different laying years on the C value, and the residual chlorine decay coefficient K value.
Water supply network modeling is widely used, and can solve problems that were difficult to solve in the past for enterprises in the operation and management of pipe networks, such as: formulate plans for water supply plants, optimize scheduling design schemes; formulate renewal and reconstruction plans for water pumps and pipelines; understand pipe networks System performance; evaluate the water supply capacity of the existing pipe network; solve pipe network abnormalities (such as valve closure, etc.); reduce pump operation costs; emergency measures for pipe network accidents (such as pipe bursts); pipe network water quality analysis; urban pipe network district water supply planning and design. Water supply network modeling is not done overnight, it needs to be continuously developed, updated and improved.
According to the modeling method, the pipe network model can be divided into the following three types of models: ① pipe network macro model; ② pipe network simplified model; ③ pipe network micro model.
The macro model of the pipe network is based on the premise that the flow of the pipe network obeys the "proportional load", and the basic idea of "black box theory" is applied to directly establish the relationship between the "input volume" and "output volume" of the water supply system. Usually, the water supply pressure and flow rate of the water plant are used as the "input quantity", and the pressure at the pressure monitoring point is the "output quantity". The macro model of the pipe network is based on a large number of measured data bits of the water supply pressure and flow of the water plant, the water level of the water tower and the pressure of the pipe network monitoring point, and an empirical mathematical expression established by applying statistical mathematics methods, thereby improving the calculation speed. However, it requires enough pressure monitoring points, and requires sufficient measured data and accuracy. If necessary, it must refer to the simulation calculation results of the microscopic model for correction. The macro model of the pipe network has achieved good results when used in the "proportional load" pipe network system. The so-called "proportional load" means that the total water consumption of the pipe network system and the flow of each node change in the same proportion at different times of the day. The water supply system of most cities in my country supplies industrial and domestic water at the same time, which does not meet the "proportional load" condition, so the concept of "period macro model" is proposed. That is to say, a day is divided into several time periods, so that the "proportional load" condition is basically satisfied in each time period, and then a macroscopic model of the water supply system is established for each time period. The macro model cannot obtain the working condition parameters of pipe sections and nodes, and is mostly used for water supply system scheduling modeling, and is not suitable for water supply system construction, reconstruction and expansion modeling.
The simplified model of the pipe network is to build a simplified model of the pipe network by selecting only the water delivery pipes with relatively large diameters, ignoring the water distribution pipe network with relatively small diameters. Since there is often no clear distinction between the water delivery pipeline network and the water distribution pipeline network in the water supply pipeline network, how to determine the water delivery pipeline section is a problem that needs to be carefully considered. In addition, in order to ensure the continuity of the pipe network, or to retain more important connecting water pipes, it may be necessary to select some pipe sections with smaller diameters to supplement. Due to inappropriate design or incomplete information, the diameter of some important connecting water pipes may be too small, which should be carefully analyzed and dealt with. Establishing a simplified pipe network model can indeed greatly reduce the workload of modeling, but some problems in the simplified model must be considered: ①Whether the simplified model of the pipe network is equivalent to the actual pipe network, and how much error will be generated after simplification is worthy of further study and discussion The problem. ②Guaranteeing the continuity of the pipe network is only the minimum necessary condition for the simulation calculation, but not a sufficient condition. It is necessary to carefully consider how to add smaller pipe sections while further simplifying. ③ Determining the water supply range of each node is a very complicated and meticulous work. Most of the user's access pipes are on the simplified pipe sections with a diameter of less than 300mm, so their positions in the pipe network cannot be displayed. In addition, some method must be used to determine the corresponding relationship between each user and its water supply node, and a corresponding program must be prepared to convert the user's water consumption to its water supply node. However, in some special cases, the water supply scope of the node may be different from the normal situation, and it is difficult to deal with it.
The microscopic model of the pipe network includes all elements of the pipe network (pipe sections, valves, water pumps, etc.), and is a model established without any simplification. Its most obvious advantage is that it directly applies the complete and detailed pipe network information database, including all the information of the pipe network for modeling, and its calculation results can obtain all the information of all nodes and pipe sections; its disadvantage is that the calculation workload is large, It takes a long time to calculate and takes up a lot of computer memory. Due to the development of computer technology, the calculation speed has been greatly improved, coupled with the further improvement of the calculation method, the simulation calculation of the operating conditions of the pipeline network can be carried out directly by using the microscopic model of the pipeline network without any simplification.
The establishment of a water supply pipe network model is a highly systematic work, and the collection and sorting of early data has a great impact on the quality of the model. The technical process of pipe network modeling is shown in Figure 1-1.
1-1 Technical flow chart of water supply network modeling
2.1 Model database construction
2.1.1 Basic data collection and processing
Before modeling the water supply network, it is necessary to investigate the actual situation of the project to be modeled, formulate a detailed and feasible work plan based on the investigation, and carefully evaluate the workload and difficulty of the entire project. The basic data survey scope of the project is as follows:
1. Pipe network
It mainly includes: the completeness of the existing pipe network data and the basic situation of the pipe network in the district; the update status and storage method of the current pipeline data; the water supply service situation in each district; Mainly control the position and opening of the valve, the number of flowmeters, the number of main accessories, the detailed location and the status of use), the buried depth of the pipeline, the age of laying, the completeness of the pipe material and other information; the use of the original water supply geographic information system and updates, as well as related hardware configuration and personnel.
2. Waterworks
It mainly includes: the water sources of each water plant; the design and operation data of each water plant; the basic data of the second pumping station.
3. User
It mainly includes: the basic situation of the software used in the water supply business charging system; the classification of charging users, the basic situation of IC card charging; the storage format and openness of the user water consumption database; the consistency of the user number of the water supply business charging system and the SCADA system.
4. Other
It mainly includes: the number of fire hydrants and the feasibility of on-site measurement; the exact position of fire hydrants and the diameter of the pipe section where they are located; the manpower cooperation of the water supply company and the provision of tools.
After the investigation of the basic data of the project, the modeling data needs to be collected and processed. Depending on the modeling type and application direction, the data collected for modeling is also different. The specific collected data can be seen in Table 3-1.
After the basic data of the pipeline network is collected, the collected data should be processed. The data processing includes the collation, evaluation and analysis of the data. The principles and contents of the collation, evaluation and analysis are as follows:
(1) Data representativeness;
(2) Timeliness of data;
(3) Data integrity;
(4) Data distribution;
(5) data defects;
(6) Data accuracy;
(7) There are problems;
(8) SOLUTION.
2.2.2 Model topology establishment
There are three main ways to establish the pipe network topology:
(1) Manual input to establish topology: For pipe network modeling projects that only have paper drawings, digitizers can be used to digitally input water supply facilities into the modeling software to form pipe network graphics with point-line structures;
(2) The establishment of the topology structure is realized by data conversion between different files: in the pipe network project with electronic CAD drawings, the pipe network data in the CAD drawings can be processed and extracted into the relevant modeling software, but in the Some cases also require some manual editing;
(3) Directly use various files to establish the topology of the pipe network: directly import the vectorized graphic data and attribute data of water supply facilities in the GIS system according to a certain file format, and form the computer water supply system pipe in the pipe network model software web graphics.
In the process of establishing the topology of the pipe network, the actual pipe network graph must be "simplified with micro-errors" according to the modeling requirements. The principle of "minor error simplification" is as follows:
(1) Simplify branched tubes. In general, branch pipes are simplified, and their user water consumption is calculated to the ring pipe nodes connected to it. If there is a large user on the branch pipe, the branch pipe will be reserved to facilitate the 24h real-time water consumption curve investigation of the large user.
(2) Simplify the "T" shaped connecting pipe. When the diameter of the connecting two pipes "T" is the same as that of one of the pipes, "T" will be simplified; if the diameter of the connecting two pipes "T" is different from the diameter of the two pipes, "T" will be retained ; If there is a normal operating valve on the "T", the "T" will be retained.
(3) If the water pipeline is not a branch pipe, no matter how much the diameter is, it cannot be simplified.
(4) The pipe network structure cannot be manually decomposed, and parallel pipelines cannot be merged.
(5) Multiple pipes are connected to the same pipe, and the distance between nodes is less than 2m, they can be merged into one node, and the local head loss is converted into head loss along the pipeline.
(6) The water head loss caused by pipeline turning or "minor error simplification" is converted into the water head loss along the pipeline during model verification.
When establishing a pipe network topology, attention should be paid to the selection of pipe network nodes, and the selection of nodes should follow the following principles:
(1) Nodes are set at both ends of the valve and the pump;
(2) Reservoir or water tower setting nodes;
(3) Nodes are set at the intersection of pipelines;
(4) Nodes are set at the variable diameter of the pipeline;
(5) Nodes are set at the joints of pipelines buried in different ages (generally more than 5 years old);
(6) Nodes are set at the joints of pipes with different pipe materials;
(7) Set a separate water consumption node at the location of the large user;
(8) Nodes are set at the on-site test points.
2.2.3 Pipe network simplification
In the pipeline network modeling, due to the complexity of the pipeline network, there are many crossing and branch pipeline networks. In order to ensure the clarity of the pipeline network topology, we generally simplify according to the following principles.
(1) Delete the secondary pipeline and keep the main pipeline and main pipeline;
(2) Merging of similar intersections to reduce the number of pipelines;
(3) Delete the fully open valve, retain the regulating valve, pressure reducing valve, etc.;
(4) Hydraulic equivalent combination of series and parallel pipelines;
(5) The large system is split into multiple small systems and calculated separately.
The final result provides the basis for partitioning by calculation,
2.2.4 Node traffic distribution
The distribution of pipeline network node flow is a key step in establishing the hydraulic adjustment model of the pipeline network, which is related to the accuracy of the model. The calculation of node flow is based on the user's water consumption, which may be the most important, the most variable and the most difficult to determine accurately in the dynamic information of the pipeline network.
The research shows that the change of water consumption by users has the following rules: ① Changes in a one-day cycle can be represented by an hourly change curve; ② changes in a one-week cycle can be represented by a weekly change curve. It mainly reflects the difference between working days and rest days; ③The changes taking January as the cycle can be represented by the monthly change curve; ④The changes taking the year as the cycle can be represented by the annual change curve; The improvement of biochemical conditions showed a trend of increasing water consumption.
In addition to the above determined factors, the influence of random factors such as legal holidays and weather changes on the law of water consumption changes should also be considered. In addition, the changing patterns of water consumption of different types of users are also different. Generally, users are divided into various types such as industries, residents, and institutions, and their water consumption changes are measured to determine their water consumption change curves. When calculating node traffic, users can be divided into two categories: centralized users and decentralized users. The water consumption of large centralized users can be converted to the corresponding nodes according to the location of their access pipes, and the water consumption of scattered users can be converted to the corresponding nodes according to the length of their pipe segments and the respective conditions of users on the pipe segments.
The selection of centralized large users should be based on the scale of the pipeline network, the size of the city and the specific conditions. For example, users with water consumption greater than 15,000m³/month are extra-large users and can be used as a water consumption node; users with water consumption greater than 5,000m³/month are large users and can be used as a water consumption node. Generally, the sum of the water consumption of extra-large users and large users is about 40% to 70% of the total water consumption of the pipe network. Water consumption measurement is a very important but extremely complicated task that requires a lot of manpower. The time-varying curves of water consumption of various users can be obtained through on-site measurements as the basis for calculating node flow.
There are two main schemes for the calculation process of converting user water consumption into node flow, as shown in Figure 2-1 and Figure 2-2.
Figure 2-1 Calculation process for converting user water consumption into node flow (Scheme 1)
Comparing the two schemes, scheme 1 converts the flow rate of the pipe section according to the ratio of the pipeline unit length to the flow rate method. This scheme needs to establish the relationship between each scattered small user and the water supply node, and needs to determine the situation of small users on each pipe section separately, so that Distribute flow by pipe length. This processing method is relatively simple but crude. Option 2 is to determine the node flow according to the water supply area method. During the calculation process, the water supply range of each node needs to be determined. All the water consumption of users within this range will be converted to the node according to the change curve, so it is necessary to establish a node and charge Relationship between bills, determined for each user
Figure 2-2 Calculation process for converting user water consumption into node flow (Scheme 2)
(including all users except very large users), so it is more accurate. But for a large pipe network, usually there may be tens of thousands of users, each user must determine its water supply node is very time-consuming.
In determining the change curve of user water consumption, it mainly includes two aspects: ① analyze and summarize the user water consumption charge records for several consecutive months (at least 12 months), and determine the annual and monthly change curve of water consumption; Reasonably classify and actually measure the daily and time-varying curves of each type of user.