IoT ecosystems for local energy systems

Posted on

< Project Overview >

Introduction

One of IUK’s four Large Scale Demonstrators for a future energy system is the Local Energy Oxfordshire (LEO) project which is set to introduce a novel local flexibility market in Oxfordshire. This market requires a system that can measure, process and validate trades. DCC will be focussing on the communications link and the University of Oxford will focus on the technologies required for validation.

Project aims

The aim of the project is to use IoT in order to validate physical execution of trades in a local flexibility market. The main barriers are related to technology maturity and lack of operating models. Currently there is no incentive for the stakeholders to collaborate and no means to gauge this activity. This is important to overcome since local energy systems and markets will gain momentum as we progress towards a net zero future. The flexibility market operator and trading parties will benefit from a reliable market. The community and environment will benefit from a local flexibility market which can accommodate a higher share of renewables.

What was done?

The main components of this project are: Data acquisition, Communication and Analytics.

A measurement module was developed to be installed at the site. This consists of a three- phase power meter fitted with a raspberry pi. The analytics package tested various types of machine learning algorithms to process data. DCC contributed to the project with feedback on the communications link required.

Results

The results reveal that the performances of both supervised and unsupervised algorithms are comparable. Given this information, the decision would be based on other factors such as computational capacity and the communication data transfer limit available at our disposal. Generally, supervised learning methods involving neural networks require more intense computation.

As the prediction horizon and the amount of training data increases, we will reach the limits of the analytics that can be deployed globally making unsupervised the more suitable alternative.

Impact

Project outlines the requirements for a baselining system. It lays the groundwork for a trial involving a medium sized fleet of mixed local energy assets. This project has been presented in a technical workshop hosted by the University of Oxford.

Next steps

Deploy technology on more assets and check effectiveness. Identify requirements for tight integration of measurement, analytics and communication. Test communications options that DCC can provide for economic and technical viability.

Lessons learned

We have been able to refine trial design and protocols for a larger trial of this technology. We have been able to develop prototypes of software and hardware platforms required for baselining and thus determine suitable tools for future roll outs.

COVID restrictions hampered progress and caused delays. Anticipating such delays would have allowed us to accelerate iterations of prototype testing.

The project would have benefitted from access to a mature flexibility market. Simultaneous progress of the baselining and market/asset-fleet development would have allowed us to test the technology on the real world earlier.

Access to deliverables, resources and media content

Components and error histograms are below. For more detail, please contact Prof Malcolm McCulloch via Mallory.newman@eng.ox.ac.uk

Measurement module:

Analytics error histograms

What has Pitch-In done for you?

Pitch-In has laid the foundation on which future baselining technologies will thrive. It has highlighted that communication limits and technology options can have significant impacts on design and development.

Project Lead

Professor Malcolm McCulloch – University of Oxford