Digital Twin LLC (DT) was founded in 2019 under biggest russina venchure investment fund to develop and promote the Digital Twin methodology and software, which should have a significant impact on increasing the efficiency, reliability and security of social, economical & technological systems.
The company’s specialists are engineers, programmists, mathematicians and physicists, former leading experts of IBM, Transneft and Sberbank.
Petroleum Engineering & Development Company transfer Crude Oil from facilities in Province to Port in Province with the capacity of 100 bbl/day via approximately 100 km pipeline
The objective of request for proposal is “Development of software for pipeline leak detection” by hybrid Mass balance method & Negative Wave Pressure Method or Real Time Transient Model (RTTM/Extended RTTM) depending on our decision.
Major concern is Real Time Monitoring 1000 km Pipeline with large diameter (42”) and establishing Virtual Pipeline1 to calculate how the flow in the pipeline should be if there is no leak.
To answer on that objectives & concerns Digital Twin made the proposal for software development with installation additional infrasonic devices based on some assumtions:
The achievements in the field of rapid detection and localization of leaks in oil pipelines (leak detection systems - LDS) today look something like this:
The minimum detectable leak is 1.8% of the pipeline flow rate for stationary pumping mode and 6% for non-stationary mode. For the oil pipeline in question (see Initial data ) losses are equal to 750 barrels per hour (~$45,000 / hour) and 2,500 barrels per hour (~$150,000 / hour).
The minimum leak detection time is 30 minutes.
The number of false positives leading to the pipeline shutdown and the corresponding losses is limited (in the some technical requirements for oil transportation) to 5 cases per year
In general, the results achieved by common technologies to be quite modest. In our opinion, this is due to the fact that in most LDS measures and analyzes the parameters of the flow of the pumped product. This approache creates the following difficulties:
Measurements of the flow characteristics can significantly depend on the properties of the pumped product, which have a significant natural dispersion.
In general, a hydraulic calculation that takes into account transients, and provides the required accuracy between minimizing leakage and eliminating false triggering, is too complex.
Many technological processes (starting / stopping of main / back-up pumping units, triggering of pressure limiting systems - PLS, beginning / ending / changing of pumping capacity by oil companies, changing of pump unit rotation speed, etc.) can lead to false alarms.
Summarizing .2, we can say that attempts to directly identify the anomalous “behavior” of the flow corresponding to leaks will always have limited accuracy due to the natural irregularity of the flow itself. Methods (for example, Negative Wave Pressure) that analyze the dynamic phenomena in the flow of the transported product, which are characteristic of leaks, looks for us more promising.
We propose to go even further: to analyze not the flow dynamics, but the dynamic behavior of the pipe in the event of leaks. The fact is that when a leak occurs, there are dynamic transverse loads on the pipeline, which can not be confused with anything (except that with the arrival of a bulldozer). These loads should lead to vibrations that can be fixed. Below we give an example of wave propagation at leakage.
The next step was to simulate the behavior of the pipeline in the event of a leak of 0.1 % of the nominal flow rate. Modeling includes:
building a 7D geotechnical model see inlet from the pump station and outlet to the pump station.
the solving of the mechanical dynamic problem, which takes into account the joint operation of the pipeline and the base with boundary conditions in the form of displacements or forces (in our case, the forces arising from the formation of a leak). As a result of solving dynamic problems (for high-loaded and low-loaded pipeline sections), the distributions of displacements, rotation angles, siols, and moments in each element and any section of the pipeline are obtained.
The distributions of movements at different frequencies (in the range from 40 to 10,000 Hz) show (see the outlet from the oil pump station and the inlet to ) that for the pipeline under consideration, a leak of 0.1 % leads to noticeable movements in the zone of ± 1 km (up to 140 mm at the NPS outlet, up to 40 mm at the inlet), which can be registered by the infrasonic device.
Digital Twin offers the integrater technology for calculating the dynamics of pipelines based on software & hardware tools for full-scale measurement of such vibrations, which can be implemented (after adaptation and calibration) as a leak detection system.
We find it very interesting to jointly solve connected thermohydraulic and mechanical dynamic problems for pipeline systems.
The technical solution based on 3 technologies developed by Digital Twin LLC:
The software supposes 3 modes of calculations.
Due to pipeline structure is continuously subjected to a complex of dynamic loads from: + environment (temperature, pressure, movement of soil and foundations), + working loads (pumping oil), + own weight and geometry, + leaks due to natural and unauthorized damage to the structure.
Digital Twin LLC offer to equip the pipeline additional ultraasonic measurement devices as shown on Diagram 1, which allow the collection of data on the dynamic loads tested by the pipeline structure.
Diagram 1. Schematic diagram of the placement of infrasound monitoring devices
The photo of infrasonic measurement device presented on Figure 1 and let the leakage system enrich data flow by mechanical dynamic characteristics and parameters of the pipeline.
Figure 1. Device photo
With the help of accelerometers and gyroscopes built into the infrasonic device, linear accelerations along 3 orthogonal axes and angular velocities along 3 orthogonal axes are measured. The measured values are recorded 400 times per second.
The operating time of the device is:
The measurement results are processed by digital filters of low and high frequency. Using numerical integration, linear displacements are determined by accelerations, and rotation angles are determined by angular velocities. Then, using the Laplace transform, the frequency response of the movements and the angles of rotation of the controlled pipeline points (where device installed).
Table 1. Technical parameters and characteristics of the infrasonic test device
| № | Parameter name | Value |
|---|---|---|
| 1 | Overall dimensions, mm | 112 х 48 х 27 |
| 2 | Mass, kg | 0.25 |
| 3 | Operating temperature, °C | from -20 till 150 °C |
| 4 | Acceleration sensor (3 axes) | MEMS accelerometer LSM303C@ST |
| 5 | Limits of linear acceleration measurement, m / s2 | from -20 to 20 |
| 6 | Angular velocity sensor (3 axes) | MEMS gyroscope |
| L3GD20H@ST | ||
| 7 | Limits of measurement of angular velocity rad/s | from -250 to 250 |
| 8 | Sensor readings recording frequency, Hz | 400 |
| 9 | Memory capacity, Gbit | 2500 |
| 10 | Maximum battery life, h | 30 |
Interpretation of the data from the device allows you to calculate the actual loads acting on the pipeline, determine the deadlines and conditions for the safe operation of the pipeline and its individual elements.
The onboard hardware and software complex of the portable device for measuring the dynamic characteristics and parameters of the pipeline provides the following operating modes:
On-board electronics provide registration and storage of the following information:
As a storage medium, non-volatile “Flash-memory” type drives with a total capacity of at least 32 Gbit are used.
The on-board software provides:
The embedded software can be installed cross-platform on Windows XP operating systems./7/8/8.1/10 and Windows Server (2003/2008/2012/2016), having a configuration not already as follows:
The standard service life is 6 years from the date of commissioning of the device for measuring the dynamic characteristics and parameters of the pipeline in operation or the operating time of 1200 charge-discharge cycles.
The product is made of aluminum carbide materials and is equipped with a thermal protection platform and a magnet for attachment to metal surfaces.
The device is transported from the supplier to the consumer and stored in the manufacturer’s packaging, subject to the conditions provided for electronic products.
Before storage, the portable infrasonic device must be preserved together with a set of auxiliary equipment for the product, a set of spare parts, a set of tools and accessories in accordance with the Operating Instructions. Products of these components that do not have standard packaging must be packed in packaging containers that protect the components from damage during transportation and protect them from external influences.
A digital twin is a digital representation of specific pipeline from Gurreh facilities in Bushehr Province to Jask Port in Hormozgan Province. The implementation of the pipeline digital twin is an encapsulated software object or model that mirrors pipeline processes, pipeline construction & natural environment of pipeline.
The delivery of pipeline digital twin includes:
Datasets package includes but not limited to the next deliveries:
Complex of program modules package includes but not limited to the next modules:
Documentation package includes but not limited to the next documents:
The initial data set for Digital Twin is the technological parameters. Our estimation for this looks like below.
Table 2. Technological design of the pipeline
| № | Parameter (indicator) | unit | value |
|---|---|---|---|
| 1 | 2 | 3 | 4 |
| 1 | Acceleration, g | m / s2 | 9,81 |
| 2 | Pipeline | ||
| 3 | Transported product-oil | ||
| 4 | Product Density | kg / m3 | 810 |
| 5 | Padding | ||
| 6 | Performance, Q | bbl / day | 1000000 |
| 7 | l / day | 158988000 | |
| 8 | m3 / day | 158988 | |
| 9 | t / day | 128780 | |
| 10 | mln t / year | 45,07 | |
| 11 | m3 / s | 1,84 | |
| 12 | m3 / hour | 6625 | |
| 13 | Flow rate (flow rate): | m / s | 2,14 |
| 14 | pressure: | MPa | |
| 15 | - inlet (pmax) | MPa | 6,3 |
| 16 | - outlet (pmin) | MPa | 0,4 |
| 17 | Pressure (oil column height): | m | |
| 18 | - inlet (hmax) | m | 793 |
| 19 | - outlet (hmin) | m | 50 |
| 20 | Pipe | ||
| 21 | Diameter | inch | 42 |
| 22 | mm | 1067 | |
| 23 | Wall thickness: | mm | |
| 24 | - min | mm | 10,3 |
| 25 | - max | mm | 15,9 |
| 26 | - calculated (highest) | mm | 14,72 |
| 27 | - by API 5L | inch | 0,625 |
| 28 | - calculated (lowest) | mm | 9,91 |
| 29 | - by API 5L | inch | 0,406 |
| 30 | Passage area | mm2 | 859605 |
| 31 | m2 | 0,8596 | |
| 32 | Steel grade | X70 | |
| 33 | Modulus of elasticity | MPa | 203000 |
| 34 | Ultimate strength | MPa | 570 |
| 35 | Yield | MPa | 485 |
| 36 | Strength Design resistance | MPa | 255 |
| 37 | Leak | ||
| 38 | Leak rate: | ||
| 39 | - highest | m / s | 125 |
| 40 | - lowest | m / s | 31 |
| 41 | Hole: | ||
| 42 | - diameter | mm | 10 |
| 43 | - area | mm2 | 78,5 |
| 44 | m2 | 0,0000785 | |
| 45 | Leak loss: | ||
| 46 | - largest | m3 / s | 0,009796 |
| 47 | - smallest | m3 / s | 0,002468 |
| 48 | - largest | m3 / h | 35,3 |
| 49 | - smallest | m3 / h | 8,9 |
| 50 | - largest (as a % of performance) | % | 0,532331 |
| 51 | - smallest (as a % of performance) | % | 0,134135 |
| 52 | The force acting on the pipe: | ||
| 53 | - from the pressure (highest) | N | 494,8 |
| 54 | - from the pressure (lowest) | N | 31,4 |
| 55 | - reactive (highest) | N | 989,6 |
| 56 | - reactive (lowest) | N | 62,8 |
| 57 | Detection | ||
| 58 | Losses due to recorded leakage: | m3 / h | 6,62 |
| 59 | m3 / s | 0,00184 | |
| 60 | - as a % of productivity | % | 0,1 |
| 61 | Hole area (at pmax) | mm2 | 14,8 |
| 62 | Hole diameter (at pmax) | mm | 4,3 |
| 63 | Hole area (at pmin) | mm2 | 58,6 |
| 64 | Hole diameter (at pmax) | mm | 8,6 |
| 65 | Force acting on the pipe: | ||
| 66 | - from pressure (largest) | N | 92,9 |
| 67 | - from pressure (smallest) | N | 23,4 |
| 68 | - reactive (largest) | N | 185,9 |
| 69 | - reactive (smallest) | N | 46,8 |
| 70 | - total (largest) | N | 278,8 |
| 71 | - total (smallest) | N | 70,3 |
| 72 | The amplitude of infrasonic vibrations (in the zone of ± 1 km) | ||
| 73 | Pressure-6.3 MPa, pipe-42" x 0.625" (outlet with NPS) | mm | <140 |
| 74 | Pressure-0.4 MPa, pipe-42" x 0.406" (PS inlet) | mm | <40 |
The log of all pipeline’s elements will be inludes in package of Data Twin delivery.
Here are examples of 2 logs of pipeline elements at the outlet and inlet to the oil pumping station, indicating:
To monitor vibrations at each section of the pipeline presented in the log - 1 infrasonic device is required.
Log of pipeline elements are the basis for creating the spatial pipeline design - the foundation of the pipeline digital twin.
Leak detection functionality Digital Twin based on precalculated patterns by “Frequency response processor” module (a.) which have to be updated (calibrated) during experimental stage (7.) of operation.
A preliminary calculation (on the theoretical design of the pipeline) shows that to ensure high accuracy and speed of leak detection, it is necessary to additionally equip the pipeline with infrasonic control devices at a distance of 1.3 km (outlet oil pumping station) - 1.8 km (inlet the station).
Thus, for the entire pipeline with a length of 1000 kilometers, it is advisable to install about 625 devices, and in total, taking into account the stock - 725 pieces.
Digital Twin technology included in the proposal allows calculate and visualize as actual place & volume of leak detection, as well as prediction the place & time for possible failuries and leakages.
The most vulnerable points, stress concentrations and residual technical resource for an example of a separate pipeline is presented below
back to 2.1.Methodology descriptionPipeline leak detection software with embedded pipeline Digital Twin as the final delivery consist of the next software modules:
| № | Component | Functionality |
|---|---|---|
| a | Frequency response processor | Calculation of the actual amplitude-frequency characteristics of pipeline vibrations based on the design documentation and current readings of measuring instruments: movements and angles of rotation of the pipeline in the places where the device is installed. |
| b | Pipeline mechanical dynamic solver | Distribution of movements and forces occurring in the pipeline elements. Calculation of the dynamic stress-strain state, including in areas of high stress concentration caused by both structural features (turns and branches of the pipeline, supports, welded joints, etc.) and defects (corrosion damage, cracks, weld defects, etc.) identified during previous surveys by non-destructive testing methods. |
| c | Preventive maintanance planner | Identification of malfunctions (deviations, defects, damages) that lead to a limitation of the safe operation period, and the causes of their occurrence. Calculation of the remaining service life and determination of the terms of safe operation of the pipeline as a whole, and its elements (pipes, connecting parts, welded joints), including: |
| d | Leakage detection clarification | Obtaining flow characteristics and pipeline movements from measuring devices that exceed the permissible (pre-calculated in module a.) value corridors. Evaluation of the probability of a false alarm, taking into account the characteristics of strength and durability (pre-calculated criteria in the software module b).) Output of a signal about the place and time of the leak occurrence. |
| e | Orchestrator for “on_demand” operation | Customizable jobs for extraxt-transformation-calculation-loading (ETL) operations from: - systems for measuring oil flow characteristics, - devices for measuring mechanical movements (vibrations) of the pipeline structure, to software module d. Leakage detection clarification with: - pre-calculated frequency response - possible places of occurrence of leaks (taking into account the operating modes of the pipeline and external environmental influences) to software module f. for Digital Twin 7D visualization |
| f | Digital Twin 7D visualisation & alarm system | 3-dimensional representation of the pipeline elements, indicating the probability of leaks and the actual leaks detected (including their geographical location, probable time and size) |
*Table 3. Functionality software description*
Diagram 2. Software component design
Modules a., b., c. is delivered out of the box with some callibrations, on the basis of a license agreement as non-exclusive rights.
Modules d., e., f. will be developed taking into account the specific of the pipeline system and its environment, on the basis of a license agreement with exclusive rights.
The work under the leakedge system software development is performed in 7 stages:
Supply of software and measuring equipment
1.1. Software delivery (Licenses + Modules)
1.2. Delivery of infrasound monitoring devices (750 devices)
Development of the digital twin of the pipeline
2.1. Creating logs of pipeline elements, connections, supports and bases
2.2. Calculations of hydraulic modes & mechanical loads 2.3. Visualization of the pipeline “as designed”, “as built”
Installation of infrasonic devices & connection with flowrate devices
3.1. Setting up and charging the devices
3.2. Connecting the devices to the current power supply network
3.3. Connecting the devices to an active data network (LTE or WIFI or FiberOptic)
3.4. Commissioning works
3.5. Analyse, setup & test connectivity with SCADA
Installation and calibration of software (modules a., b., c.) for batch processing of data on the technical condition of the pipeline
4.1. Determination of dynamic loads acting on the pipeline
4.2. Allocation of pipeline zones, the stress-strain state of which exceeds the standard
4.3. Allocation of pipeline sections with reduced rigidity
4.4. Calculation of the remaining service life and determination of the terms of safe operation of the pipeline as a whole, and its elements (pipes, connecting parts, welded joints)
Development and installation of software for real-time leak detection, visualization and orchestration of calculations (d., e., f.)
5.1. Functionality & perfomance testing
5.2. Connection to the existing data transmission network of the customer
Preparation of technical documentation
Field tests, transfer software & digital twin to production phase
The total project duration takes 9 months:
| № | Work | Results | Calendar days from contract signature | |
|---|---|---|---|---|
| Start | End | |||
| 1. | Supply of software and measuring equipment, including: | |||
| 1.1. | Software delivery | The licences to use the software modules a., b., c. | 1 | 10 |
| 1.2. | Delivery of infrasound monitoring devices | A set of devices installed on the pipeline surfaces prepared by the customer | 1 | 90 |
| 2 | Development of the geotechnical model of the digital twin | Digital twindata sets in csv. | 1 | 60 |
| 3 | Installation of infrasonic devices and connection with flow metersdevices | The leakage software could collect data from SCADA system & infrasonic devices in real-time & on-demand mode | 100 | 160 |
| 4 | Installation and calibration of out-of-the-box software | 3 modules packed in a docker container, configured for current pipeline | 11 | 71 |
| 5 | Development and installation of new modules | 3 tested software modules on R packaged into docker-container | 10 | 135 |
| 6 | Preparation of technical documentation | The boundle of users & administration guides | 160 | 165 |
| 7 | Field tests | Test reports. Confirmation of service level. | 180 | 270 |
| Total | 270 |
included as a part of Digital Twin concept↩