The electricity sector is witnessing an increasing number of grid connection applications year by year. As at the end of April 2024, in GB, the connection queue stood at 712GW[1], 48GW being demand and 664GW generation and storage. While this exceeds GB’s energy needs for net zero, there is an acknowledgment of the need for significant network reinforcements; while the on-going connection reforms continues to progress, which have shown positive outcomes with the coordinated efforts among stakeholders.
Ensuring consistency and accuracy in assessments has become more challenging due to adapting to various connection types, and technologies. The situation requires better efficiency, as further delays in getting connected to the grid will dampen progress towards the net zero.
The Electricity System Operator (ESO) and Energy Networks Association’s (ENA) have started a shorter-term tactical improvement to the connection process (both transmission and distribution) within the current industry setup. The current connection framework emphasizes on the “First ready First connected” approach, moving away from “First to apply First served”.
AI-Driven Solutions for Connection Application Process
Recognizing the need for innovation, the industry is adopting artificial intelligence (AI) as an enabler to accelerate the grid connection application processing. Energy Networks Association’s (ENA), for example, recently launched ENA Connect Direct platform to reduce waiting times by leveraging AI to provide faster approvals for routine applications. The initial reports claim that the platform is showing promising results having already processed applications from 1,000 installers in its first month of operation[2].
For low-voltage connections, the outcomes have been favorable, but addressing the greater complexity involved in medium and high voltage connections remains a challenge that calls for more elaborate solutions. Coordination among the ESO, network owners and the connection applicants can possibly be enhanced with greater transparency as well as process automation enabled by an AI.
At PSC, we have identified three touchpoints along the connection process where AI could make an impact and produce a significant improvement; application submission, system assessment and issuance of connection offers. Later in this briefing, we will touch on the implementation strategy and next steps.
Touchpoints for an AI-powered connection process
1. Application submission
Reducing manual intervention in processing applications means lessening the time taken while ensuring accuracy and completeness of each submission. Ideally, an application form evaluation component would categorize applications based on their characteristics and predict potential outcomes of success and failure. To achieve this, the component would require training a machine learning model on historical data from previous applications to recognize patterns and issues in submissions. The outcome of the AI-reviewed applications will guide the next steps of the process allowing timely or immediate feedback to the applicants.
2. System assessment
The studies necessary for evaluating grid connections involve detailed analysis of factors such as fault levels, voltage stability, and power flow to ensure that new connections do not compromise the capacity, reliability and stability of the grid. A critical one is transient stability analysis, which assesses the system ability to maintain synchronism when subjected to severe disturbances. The increasing penetration of renewable energy sources, make these more challenging as they don’t inherently contribute to system inertia, and the outcome frequently requires additional assets to be installed to meet the grid security standards.
While it is an open question whether an AI agent would be able to perform power system studies by itself in the current state-of-the-art, an automated assistant can be expedite the studies pipeline by structuring and pre-analyzing the input data for the specialists running the studies. A team of Power Systems experts together with an AI specialist, could train a model to estimate the impact of proposed connections on grid stability and performance, utilizing historical data available. This historical data must include earlier connection studies, cross-referencing between disparate studies to improve the AI model generalization capability regarding general system limitations.
3. Issuance of connection offer
In the final stage of the connection process, an AI system could be employed to enhance quality control of the connection offer. This module would focus on ensuring consistency between various components of the offer and finally crafting an offer to the applicant using the input generated from the reviewed application form and from the power system studies.
This Generative AI model, established and fine-tuned using a historical dataset of the connection offers and system assessments, would potentially be capable of recognizing patterns and discrepancies. However, the final approval and issuance of the offer would remain the responsibility of experienced professionals, ensuring that important investment decisions are subject to appropriate human oversight and expertise.
Roll-out strategy (pre and post Go-Live)
A successful implementation of a project of this size and scale is contributed by effective communication framework among the stakeholders, robust quality assurance and a blend of AI development skillsets, engineering as well as business process knowledge in electricity industry.
Achieving the project objectives requires a deep understanding of the issues and appreciation of the end-to-end connection process before moving to the next phase of problem solving and system development. Therefore, a project team with a mix of data science, business process analysis and engineering expertise with sufficient industry knowledge are essential to develop predictive models and automation scripts that align with the industry requirements and more importantly robust enough with some room of flexibility to adapt to future changes.
Quality assurance is important considering the innovative nature of a project like, this with realistic goals and expectations. A comprehensive change management strategy will ensure flexibility in adapting to project shifts, along with the adoption of agile principles for a dynamic project environment.
The proposed AI solutions must be validated before deployed in real-life. Pilot tests prior to post go-live could provide valuable performance data and help identify potential shortcomings. As part of the quality assurance process, continuous engagement with all stakeholders including decision-makers, grid operators, network owners, and targeted group of the system users will ensure consistent feedback and alignment of the proposed solution.
Most importantly, during the post go-live phase, development should not end with the project there. Continuous improvement is necessary and future changes are expected, progressing with the industry dynamics. As technologies and the regulatory environment evolve, the proposed AI systems must be future-proof, with capability to integrate additional data sources and adapting to changing operational requirements.
While AI can certainly improve efficiency and quality assurance, it’s important to remember its limitations. Specifically, AI cannot overcome the physical constraints of the power grid; it won’t enable faster connections without the necessary transmission infrastructure in place. The technology should be viewed as an enhancer of existing processes rather than a magical solution to physical limitations.
How PSC can help
It is important to recognize that this advancement requires a collaborative effort. It necessitates the union of AI experts and power system professionals to develop solutions that meet the specific and complex needs. With the right expertise, approach, and ongoing commitment to improvement, PSC can assist its clients in leveraging AI to enhance grid connection processes and to build a more responsive and efficient energy infrastructure capable of meeting the challenges of the future.
PSC bring together worldwide experience and global leaders in their field to develop innovative solutions that allow utilities and energy companies to thrive. This allows PSC to provide excellent value to our clients whilst working to deliver our vision for a sustainable power system whilst improving the quality of human lives. We combine power network experts, operational technologies experience and extensive knowledge of strategy and regulation to address challenges facing the industry.
PSC’s global energy experts will help your organization deliver innovative solutions. Please find out more about our capabilities in this area and contact us to talk about the first steps.
References
[1] 240612may-connections-delivery-board-meeting-minutes.pdf (energynetworks.org)
[2] ENA Connect Direct platform hits 1000 clients in first month (current-news.co.uk)