US-12618718-B2 - Method and system for selection and monitoring of a steam trap using deep neural network

US12618718B2US 12618718 B2US12618718 B2US 12618718B2US-12618718-B2

Abstract

The invention relates to a method and system for selecting and monitoring of a Venturi type steam trap in a steam system using deep neural network to process specific parameters of the steam system stored in memory of a data processing unit with the use of temperature detection only.

Inventors

Mateusz Śmiechowicz
Andrzej Grzebielec
Miroslaw Seredyński
Piotr Łapka
Emil Gromadzki

Assignees

ENEON SP. Z O.O.

Dates

Publication Date: 20260505
Application Date: 20220127
Priority Date: 20210128

Claims (15)

1 . A method for selecting and monitoring of a Venturi type steam trap in a steam system, comprising: measuring a current temperature of a condensate at an inlet of a steam trap by a temperature sensor placed at the inlet of the steam trap, measuring a current temperature of the condensate at an outlet of the steam trap by a temperature sensor placed at the outlet of the steam trap, optionally, measuring a current temperature at the inlet of a technological process operated by the steam system by a temperature sensor placed at the inlet of said technological process, determining, by a data processing unit having a processor and memory, current working conditions of the steam trap and optimal characteristics of the steam trap in the steam system using deep neural network designed to determine optimal characteristics of the steam trap, by: processing parameters stored in the data processing unit memory, wherein the parameters comprise: a type of steam system, steam pressure, condensate pressure, nominal mass flow rate of the condensate, nominal temperature of the condensate at the inlet of the steam trap and nominal temperature of the condensate at the outlet of the steam trap, calculating optimal characteristics of the steam trap in the steam system, wherein optimal characteristics of the steam trap in the steam system include orifice diameter of the steam trap and orifice length of the steam trap, processing of the current temperature of the condensate at the inlet of the steam trap and the current temperature of the condensate at the outlet of the steam trap to obtain current working conditions of the steam trap, optionally, processing of the current temperature at the inlet of the technological process operated by the steam system to obtain further current working conditions of the steam trap, outputting data representing current working conditions of the steam trap and optimal characteristics of the steam trap in the steam system.
2 . The method according to claim 1 , further comprising: installing in the steam system a Venturi type steam trap having an orifice diameter of the steam trap and an orifice length of the steam trap selected according to optimal characteristics determined by the data processing unit.
3 . The method according to claim 2 , further comprising signalling, by a signalling module of current working conditions of the steam trap, wherein signalling includes acoustic signalling, optical signalling, mechanical signalling and electronic message.
4 . The method according claim 1 , wherein the temperature sensor is selected from the group consisting of NTC sensor, wired sensor, wireless sensor and combinations thereof.
5 . The method according to claim 3 , wherein signalling further comprises at least one of current working conditions: the steam system is working properly, the steam trap ensures a discharge of a correct amount of the condensate, the condensate is discharged too quickly, leading to energy losses due to escape of a live steam, the condensate builds up in front of the steam trap, due to its too low discharge, the selection of the steam trap mounted in the steam system is not optimal, an error in a plurality of calculations based on a flow model is detected, a change in the operating conditions of the steam system has occurred, current efficiency of the steam trap, temperature characteristics diverging in time, incorrect time correlation between measured temperatures.
6 . The method according to claim 1 , wherein the steam system is one of the steam systems comprising: steam pipeline, pipeline heating a flowing medium, heat exchanger, liquid separator, heated tank, including mixer and batch heater, turbine, vulcanization autoclave, or system used for hygienic applications.
7 . The method according to claim 1 , wherein the method is performed in a loop, wherein in at least one iteration, the current temperature of the condensate at the inlet of the steam trap and the current temperature of the condensate at the outlet of the steam trap are recorded in memory of the processing unit and are used in at least one further iteration instead of the nominal temperature of the condensate at the inlet of the steam trap and the nominal temperature of the condensate at the outlet of the steam trap, respectively.
8 . The method according to claim 1 , wherein the steam trap is mounted in a steam system using mount types comprising: internal thread, external thread, flange connection, tri-clamp mount.
9 . A system for selecting and monitoring of a Venturi type steam trap, comprising: a Venturi type steam trap mounted in a steam system, a temperature sensor placed at an inlet of the steam trap for measuring a current temperature of a condensate at the inlet of the steam trap, a temperature sensor placed at an outlet of the steam trap for measuring a current temperature of the condensate at the outlet of the steam trap, optionally, a temperature sensor placed at the inlet of a technological process operated by the steam system for measuring a current temperature at the inlet of said technological process, a data processing unit having a processor and memory, configured to determining current working conditions of the steam trap and optimal characteristics of the steam trap in the steam system using deep neural network designed to determine optimal characteristics of the steam trap, by: processing parameters stored in the data processing unit memory, wherein the parameters comprise: a type of steam system, steam pressure, condensate pressure, nominal mass flow rate of the condensate, nominal temperature of the condensate at the inlet of the steam trap and nominal temperature of the condensate at the outlet of the steam trap, calculating optimal characteristics of the steam trap in the steam system, wherein optimal characteristics of the steam trap in the steam system include orifice diameter of the steam trap and orifice length of the steam trap, processing of the current temperature of the condensate at the inlet of the steam trap and the current temperature of the condensate at the outlet of the steam trap to obtain current working conditions of the steam trap, optionally, processing of the current temperature at the inlet of the technological process operated by the steam system to obtain further current working conditions of the steam trap, outputting data representing current working conditions of the steam trap and optimal characteristics of the steam trap in the steam system.
10 . The system according to claim 9 , further comprising a signalling module, configured to signalling of current working conditions of the steam trap, wherein signalling includes acoustic signalling, optical signalling, mechanical signalling, and electronic message.
11 . The system according to claim 9 , wherein the temperature sensor is selected from the group consisting of NTC sensor, wired sensor, wireless sensor and combinations thereof.
12 . The system according to claim 10 , wherein the signalling module is further configured to signalling at least one of the current working conditions: the steam system is working properly, the steam trap ensures a discharge of a correct amount of the condensate, the condensate is discharged too quickly, leading to energy losses due to escape of the a steam, the condensate builds up in front of the steam trap, due to its too low discharge, the selection of the steam trap mounted in the steam system is not optimal, an error in a plurality of calculations based on a flow model is detected, a change in the operating conditions of the steam system has occurred, current efficiency of the steam trap, temperature characteristics diverging in time, incorrect time correlation between measured temperatures.
13 . The system according to claim 9 , wherein the steam system is one of the steam systems comprising: steam pipeline, pipeline heating a flowing medium, heat exchanger, liquid separator, heated tank, including mixer and batch heater, turbine, vulcanization autoclave, or system used for hygienic applications.
14 . The system according to claim 9 , wherein the system is configured to operate in a loop, wherein in at least one iteration, the current temperature of the condensate at the inlet of the steam trap and the current temperature of the condensate at the outlet of the steam trap are recorded in memory of the processing unit and are used in at least one further iteration instead of the nominal temperature of the condensate at the inlet of the steam trap and the nominal temperature of the condensate at the outlet of the steam trap, respectively.
15 . The system according to claim 9 , wherein the steam trap is mounted in a steam system using mount types comprising: internal thread, external thread, flange connection, tri-clamp mount.

Description

TECHNICAL FIELD The present invention relates to a method and system that use deep neural network for selection and monitoring of a Venturi type steam trap used for draining-off liquids from steam driven process predominantly containing gases or vapours. In particular, the present invention allows identifying current working conditions of the steam trap and determining geometrical conditions of the steam trap to achieve optimal characteristics of the steam trap. BACKGROUND ART Steam traps are used in many industrial applications to remove a condensate and non-condensable gases such as air while ensuring the live steam is not wasted and is effectively used in the steam system. There are three known types of steam traps: mechanical traps, thermostatic traps, and thermodynamic traps. Mechanical traps remove condensate by mechanical properties of steam and condensate and these traps may be equipped with valves that open and close or a bucket or a float that rises and falls in relation to condensate level. Thermostatic traps remove condensate through the temperature difference of a steam and a liquid phase and sequentially open and close taking temperature drop below the saturation curve through a bimetallic element. Thermodynamic traps operate on the dynamic principals of steam and condensate and the use of Bernoulli's principle depending on the relationship between the velocity and the pressure exerted by the condensate and steam inside the steam trap. One kind of thermodynamic trap is Venturi type steam trap in a form of Venturi nozzle having an orifice creating a flash effect. Venturi type steam traps do not have mechanical parts and are thus more reliable and durable without the frequent need for a maintenance. In order to work with maximum efficiency Venturi type steam trap has to be carefully selected for the specific steam system. WO/1995/022714 A1 describes a continuous flow steam condensate removal device having an entrance end, an entrance passageway, an intermediate passageway which includes a cylindrical venturi tube adjacent an exit of the entrance passageway and a downstream cylindrical discharge-transition passageway adjacent an exit of the venturi tube. WO/2014/147381 A2 describes a condensate removal device having a condensate drainage channel with a constricted passage forming an orifice therein. A magnet is disposed upstream of the orifice to capture impurities, e.g. magnetic particles, suspended or otherwise carried in condensable gas or condensate. WO/2016/174691 A1 describes a device for the determination of optimal diameter of the orifice of a Venturi nozzle steam trap in operation in a steam flow circuit. The condensate out from the accumulator is led to the block with a rotating disk-provided with a series of orifices with varying diameters and in a predetermined range of sizes-moved by a stepper motor. For each disk placement, values of condensate level L, temperature T and pressure P are detected by means of sensors placed upstream and downstream of the disk and are transmitted to the processor to verify, for each orifice diameter and related disk position, the phase state of the fluid upstream and downstream of the disk. EP0426199 A2 describes a trap selector for selecting an optimum trap meeting various uses which is designed to sequentially input required data for selecting the trap and to execute previously memorized trap selecting programs by a microcomputer, wherein the geometrical conditions of the trap are expressed by valve flow coefficient of various kinds of traps are contained in the memory with small capacity without memorizing the discharging flow capacity for various working conditions of various traps. WO/1998/010218 A1 describes a self-contained electronic system for the monitoring and continuous surveillance of purgers, valves and installations using fluids. The system is comprised of three essential elements: a multiple sensor, an electronic analyzer and an optional receiver. The sensor measures the conductivity, the pressure and temperature of the fluid. WO/2005/034046 A1 describes a system diagnosing method, comprising the steps of: performing a trap operation diagnosis; calculating a total trap-passed steam loss amount obtained by aggregating trap-passed steam loss amounts for all the evaluation target steam traps; and generating comprehensive evaluation data. For diagnosing operational conditions of the steam traps the system a detection of surface temperature and vibration (vibration intensity in ultrasonic range) of the steam trap. WO/2016/163163 A1 describes a threshold value calculation system and a threshold value calculation method for calculating threshold values that are used to determine the state of a steam trap. The state-of-the-art methods, systems and devices for the assessment of the working conditions of Venturi type steam trap use a combination of detection methods, e.g., a combination of conductivity and temperature detection or a combination of