Study On Abnormal Parameters Correction And State Estimation Of Oilfield Water Injection Network

The purpose of establishing oilfield water injection network is to distribute water from the water injection pump station to each injection well according to the actual production needs,so as to meet the needs of injection water pressure and flow of each injection well.In the long-term operation of the water injection network,there will be serious scale,corrosion,perforation and other pathological phenomena,which will cause great changes in the friction coefficient and tube inner diameter,resulting in great errors in the simulation calculation and optimal scheduling.On the other hand,In order to monitor the operation of the entire water injection system and collect water injection data,we need to install pressure and flow sensing devices at each node of the pipe network.However,due to the limitation of investment cost and technology,the real-time information monitoring system only installs sensor devices at the main nodes.How to estimate the pressure and flow value of each node of the whole pipe network system through the detection data of the limited nodes is an urgent problem to be solved in oilfield production management.To solve above problems,first select an object to actual parameters for inversion of pathological network,node were collected in the actual pressure and flow rate as the basis,through the reasonable inversion algorithm to solve the network equivalent parameter of the relatively correct(for simplicity,all pathological parameters due to the influence of the change of the equivalent friction coefficient),provide accurate model for the simulation calculation of network data.On this basis,this paper studies an algorithm to estimate the pressure state of the whole pipe network by using the pressure flow data of some known nodes in the pipe network,and applies it to the actual pipe network state estimation.This paper mainly studies the following contents.In order to solve the problem of water injection network simulation,a mathematical model of water injection system network adjustment was established.It is proposed to optimize the mathematical model of ideal pipe network by using quasi-Newton method and least square method respectively,so as to train the best solution.Aiming at the distribution law of finite pressure test points,the mathematical model of pressure test point decision problem in pipe network system is established,and the optimal classification of pressure test points is determined by fuzzy clustering method.Aiming at the problem of modifying the ill-condition parameters of the pipe network with unknown nodal pressure,namely the inversion of the friction coefficient of pipe network elements,the objective function model for the inversion of the friction coefficient of theoilfield water injection pipe network was established.Particle swarm optimization and simulated annealing algorithm are used to solve the iterative optimization of friction coefficient inversion of multi-variable and multi-parameter pipe networks.Aiming at the problem that the real-time information monitoring system only installs the pressure and flow monitoring device at the main nodes,a mathematical model for estimating the status of oilfield water injection pipe network is established.Tabu search algorithm and simulated annealing algorithm are used to solve the model iteratively.The calculation results show that it is feasible to estimate the operation state of the pipe network based on some known node parameters.A set of oil field water injection simulation device is established,and the algorithm of network parameter correction and state estimation is applied to the device.The algorithm is modified and verified with the measured data of the simulation device,and the correctness of the algorithm is proved.