Network restrictions in IoT-based microgrids
26/05/21
< Project overview >
Introduction
The project investigated the role of IoT technologies in the control and management of microgrids (a microgrid is a localised distribution system comprising loads, generation and energy storage that can be operated autonomously or connected to a larger grid). An offshoot of this work involved contributing to the development of a particular power-converter based datalink technology.
The main collaboration has been with Shanghai Jiao Tong University (Dr Fei Gao, working on simulation of fast energy management) and internally at Oxford (RELCON project team, working on the datalink technology). Unfortunately, COVID- 19 reduced the scope for further outside collaboration and limited the amount of lab-based experimental work that was possible.
Project aims
To explore what ‘internet-based’ technologies may be able to achieve in terms of very fast energy management in microgrids. The principal focus was to use ubiquitous ethernet links to facilitate centralised and semi-decentralised approaches to energy management with cycle times of ten milliseconds or less. This contrasts with the usual energy management cycle times of several seconds or minutes.
To investigate methods of transferring data between power electronic converters over the power connections using the in-built controllability and sensing capability of the converters themselves (ie without using traditional PLC hardware such as GreenPHY).
The principal direct beneficiaries of this project will be any entity that controls a low-voltage grid, in both developed and developing countries. This technology will allow them to provide cheaper, faster and more flexible control at the end of a grid or in stand-alone micro-grids.
This is required for the fluctuation in supply from renewable energy generation and to cater for fluctuation in demand.
Amongst these stakeholders, this project aims to overcome the following barriers:
Dispositional: a lack of trust in and real potential for IoT within this stakeholder group.
Educational: a lack of familiarity with IoT to enable confident decisions regarding IoT adoption or investment.
Three mini-project final report template v1.3.1.
Technology maturity: a lack of maturity in technology and services and risks associated with adoption as a result.
What was done
Simulation and quantification of the advantages of short time-scale energy management algorithms that can be facilitated by IoT technologies (ie fast packet switched networks).
Continued the development of a ten-node microgrid hardware demonstrator based on 1Gbps wired ethernet (although this was hampered by restricted access to the laboratory due to COVID-19).
Development of an inter-converter datalink technology, including theory, hardware and software development, and hardware testing.
Results
We have developed a non-iterative algorithm for dispatching load, generation, and storage in networked microgrids.
Its non-iterative nature means that it has a defined execution time, ie that it is guaranteed to complete in a fixed time step, allowing short time-horizon management to be conducted reliably. We have trialled the algorithm experimentally on a ten-node system operating at 10 ms (100 Hz) update rates. We achieved good system balancing and energy management behaviour, although we did experience issues with stability in some cases. We theorise that this is because of variable packet delay in our network (of the order 0.5–2ms) that introduces a mismatch between command and implementation.
Unfortunately, we could not pursue this further because COVID shut down access to the lab, and Dr Fei Gao took up another post in China (although we continued to collaborate on this project).
In a separate piece of work, we developed a powerline communication (PLC) technology for exchanging information between power electronic converters. This is a novel technology in that it requires no additional hardware at either the transmitting or receiving converter – it relies entirely on the innate controllability and sensing hardware already present to operate the core power converter. We have demonstrated the concept experimentally as part of another project (EPSRC RELCON EP/R030111/1), and achieved highly reliable communication at a raw link rate of 2 kbps.
Deliverables
We intend to demonstrate a multi-converter communication network based on the new datalink technology over the course of 2021 (under the RELCON project). If the demonstration is successful, the algorithm/software could form part of the overall RELCON microgrid technology, which would have the potential for commercialisation in some form.
One of the initial aims of the project was to develop a laboratory rig that could be facility for company/academic collaborations. The Covid pandemic, and consequent laboratory closures, means this has not been achieved during this project as a practical output that would be easy for external use. However, the potential concept has been proved.
The following papers have resulted from the project:
Liao, D, Gao, F, Zhao, Y, and Rogers, D (2020) ‘A dynamic diffusion algorithm for distributed secondary control of DC microgrids’, IEEE Energy Conversion Congress and Exposition (ECCE). Detroit, MI, USA, 2020, pp. 1299–1306. doi: 10.1109/ECCE44975.2020.9235782
Conference paper (accepted): Han, R, and Rogers, D (2021) ‘Zero-additional-hardware power line communication between DC-DC power converters’, APEC 2021.
Journal paper in preparation: Han, R, and Rogers, D, ‘Zero-additional-hardware power line communication for DC-DC converters’, will be submitted to IEEE Transactions on Power Electronics.
Impact
Impact will arise from the development and validation of two new technologies: very fast energy management, and power-converter datalinks. Particularly, the datalink technology will see immediate application in the RELCON project where it will be deployed as part of a demonstration microgrid in Kenya (expected end of 2021/early 2022).
Next steps
Further work on the short time scale energy management piece is somewhat dependent on further funding and researcher availability. An MSc(res) student is currently performing some comparative analysis between different energy management schemes, including the one developed under this project. The ambition is to bring the work to a point where we can publish our findings in a top-tier journal.
The power converter datalink technology will be tested and deployed as part of the RELCON project.
Lessons learned
What went well
The project allowed us to develop our understanding of what opportunities IoT-style technologies create in power electronics-dominated microgrids.
We explored the limits of communication speed in high-end networks (ethernet at 1Gbps) when performing short time scale energy management at cycle rates of 100 Hz. This shows that such schemes are potentially feasible, but certainly require more engineering to make them robust and practical.
What could have been done better
Taking a short time-scale energy management scheme from theory to practice in the timescale of this project was ambitious. Knowing that access to the lab would be limited by COVID, we would have focussed more on the theoretical development of the technology. More focus on theory (as opposed to hardware demonstration) may also have allowed us to account variable packet delay more satisfactorily in the final implementation.
What would have been useful
It is always difficult to manage work on smaller-value but fairly long grants, as they do not allow for the allocation of a full-time researcher. Perhaps a more specific link-up with another project for the first part of the work would have been valuable – we found that the link with the RELCON project provided a context for the second part of the work.
Pitch-In’s role
Pitch-In provided an opportunity for us to develop two ideas around the active management of power converters in future microgrids. This has led to one immediately applicable technology (power converter data-links) and the ‘acquisition of experience’ around another technology (very high speed energy management).
The Pitch-In project was run adeptly by the management team (most of our contact was with Andy Gilchrist at Oxford, who was very helpful and understanding around the impact of COVID and the mitigations needed to address this.
Project lead
Professor Dan Rogers – University of Oxford