Introduction
The growth of offshore wind farm projects has increased the demand for three-core export and inter-array armored power cables of higher voltage levels. To increase the economic viability of such projects, it is crucial to reduce the subsea cable cost as much as possible. The loss calculation of such cables is traditionally performed using the IEC 60287-1-1 Standard. However, it is today widely recognized that the IEC method systematically overestimates the losses in the armor, thus leading to oversized cables. This has led Hellenic Cables to seek for a more realistic losses estimation which, apart from the development of analytical calculations and employment of finite element modelling, also includes field measurements in cable specimens. A typical three-core export cable with sheathed cores and a wire armor is seen in Figure 1.
CIGRE study committee B1 has recently constituted working group B1.64 to deal with this topic. The working group is elaborating, amongst others, on recommending a methodology for measuring the positive-sequence cable impedance, which directly reflects the total cable losses. Hellenic Cables adopts a wider and deeper analysis compared to CIGRE method, measuring not only the positive-sequence cable impedance but also the circulating losses generated from the net currents induced in screens and sheaths. This allows for a detailed allocation of the losses to the different metallic components of the cable, leading to accurate current calculations and thus optimized cable cross-sections.
Presents an overview of the cable sample configuration for loss measurements. The specimen length must be large enough to represent the actual operation of the cable and exclude end effects. In addition, it is placed in a specially designed corridor free of any adjacent metallic structures to avoid electromagnetic interference. The extreme ends of the cable sample are referred to as drive (DE) and earth (EE) ends to describe the end where the three-phase balanced power source is placed at and at the end which is grounded, respectively.
Hardware implementation via NI CompactDAQ
Hellenic Cables adopts a non-invasive, slightly intervening procedure to place at both cable ends all temperature, voltage and current sensors needed for the measurement. To have a greater degree of control over the measurement settings, a DAQ system, provided by National Instruments (NI), is used to simultaneously record all the necessary raw data. Two DAQ systems are used in total, each one per cable end. An Ethernet cable synchronizes both systems and transfers the measured data to a local PC being placed at the control room, where the subsequent data analysis is performed.
The hardware used consists of two NI CompactDAQ chassis cDAQ-9189, designed for distributed sensors measurement systems. The chassis control the timing, synchronization and data transfer between C series I/O modules and an external host. The chassis provide precise, synchronized timing over the network using Time Sensitive Network (TSN) protocol, ideal for highly distributed measurements over long distances in a daisy-chaining configuration. Each chassis is equipped with three groups of modules, depending on the signal expected as an input: NI-9247 modules are used for current measurements, being compatible with the current transformers used; NI-9239 and NI-9229 modules are used for voltage measurements, depending on the voltage level expected; NI-9216 and NI-9217 modules are used for temperature recording, being compatible with the RTD temperature sensors used.
Software implementation via NI LabVIEW
Besides the hardware used to collect raw data, a specially designed custom software, developed in NI LabVIEW by Yotta Volt, is used to calculate the cable impedance and the power losses, based on the raw data collected. The software acquires the data for visualizing in time domain and analyzes them by implementing Fast Fourier Transform analysis to detect any likely harmonics with the respective magnitude and phase. The per-unit-length positive-sequence impedance and the total cable losses, in active and reactive terms, are subsequently computed and visualized based on the measured currents and voltages in the conductors. Currents on both cable screens and sheaths are also measured and recorded. In addition, user-defined measurement duration and sampling rate are requested as input values, thus allowing for the analysis of the harmonic content. Therefore, the criterion of having a power source of low total harmonic distortion in the injected current can be evaluated and satisfied. Finally, all raw and processed data are archived in the local PC for future use. The software has also the ability to upload previous data, visualizing all results thereby enabling the direct comparison between measurements. An indicative image of the NI LabVIEW panel is illustrated in Figure.
The Challenge
To implement a measuring methodology for the accurate estimation of losses in three-core export and inter-array armored submarine power cables.
– Prior to NI technology what were your alternative solutions?
The alternative solution, prior to NI technology, was the use of commercially available wattmeter. Apart from its lower degree of control over the measurement settings and data, its accuracy is also limited at low power factors.
The Solution
| Equipment | Description | Function | Quantity |
|---|---|---|---|
| cDAQ-9189 | 8-slot TSN-Enabled Ethernet chassis | DAQ measurement system chassis for distributed sensors measurement systems | 2 |
| NI-9216 | 8 channel, 400 S/s (aggregate), 24-bit, RTD analog input module | Measurement of temperature at metallic cable components | 2 |
| NI-9217 | 4 channel, 400 S/s (aggregate), 24-bit, RTD analog input module | Measurement of temperature at metallic cable components | 2 |
| NI-9239 | 4 channel, ±10 V, 50 kS/s/ch, 24-bit, analog voltage input module | Measurement of voltage drop between metallic cable components | 5 |
| NI-9229 | 4 channel, ±60 V, 50 kS/s/ch, 24-bit, analog voltage input module | Measurement of voltage drop between metallic cable components | 1 |
| NI-9247 | 3 channel, 50 A (RMS), 50 kS/s/ch, 24-bit, analog current input module | Measurement of injected and induced currents in metallic cable components | 6 |
| NI LabVIEW | Data acquisition, numerical analysis, logging and visualization of results |
Yotta Volt offered a complete solution by delivering the NI hardware configuration and developing the tailor-made software application in NI LabVIEW that will acquire, analyze, log and visualize the measured data. The software application was developed to meet the requested functionalities and implement the measuring methodology, as discussed in several meetings and finalized in the technical quotation.
The commissioning phase was performed with a 2-day visit in Hellenic Cables factory plant in Soussaki, Corinth, Greece, which included the setup and test of the measuring system. Finally, after the commissioning phase, a debugging period was agreed to solve possible technical issues.
Results
Losses in several cable specimens have been measured with the presented methodology and the results have been compared with analytical and numerical approaches. In all cases, the measurements show very good agreement against simulations, not exceeding 5% in terms of total losses. In addition, it was reconfirmed that the IEC method systematically overestimates the losses in the armor, thus leading to oversized cables.
– Additional Benefits (time, cost benefit etc.)
The ease of use and flexibility of NI CompactDAQ, NI LabVIEW and all the existing components have enabled Hellenic Cables to create a testbed where cables of different designs can be readily measured in terms of losses. The accuracy of the measuring system has been also assessed, exhibiting a low measuring error suitable for industrial and research purposes.
Since loss measurements in three-core, wire armored cables is nothing but measurement of positive sequence resistance, the same methodology with slight modifications, including the utilization of NI DAQ systems, was later implemented to estimate the AC resistance of isolated cable conductors. Good agreement between analytical and experimental values has been achieved, thus verifying the theoretically expected skin effect factors, as well as the multi-purpose functionality of the NI DAQ system.