Learnings

Outcomes

This project conducted a detailed literature review by interviewing six leading NTO vendors and nine TSOs across Europe, USA, Asia, and Australia.  Analysis was further supplemented with desktop research on a more extensive group of global TSOs.  Current developments were captured and future plans surrounding NTO solutions. Following the literature review,  a current state assessment of prominent NTO technologies and capabilities available to date was conducted. Based on the information gathered from the vendor & TSO interview sessions, a spectrum of five categories was defined for assessing the maturity of NTO solutions, this led to development of current state technology assessment and best in class NTO solution. These investigations and analysis revealed that development of NTO solution is in early stages, hence we determined the gaps in each of the of five categories defined for assessing the maturity of NTO solutions. The estimated cost of developing the NTO solution across a multi-year program was captured, and the estimation of value to TSOs and consumers. Finally, we concluded the project by developing an implementation plan which recommends establishing a global community for the development of a usable, scalable and pragmatic NTO solution for control room applications available to all TSOs. 

The information and analysis captured during this project has been collated into a report. 
 
The outcome of this project demonstrated that NTO as a decision support tool will be an essential and daily part of power system operations, part-and-parcel of the management of a large-scale network with complex thermal, voltage and dynamic stability constraints.  

This project demonstrated that the availability of a workable NTO tool in the control room will enable control room operators to study more network scenarios and topological actions. Manual methods for determining the best topological actions are more complex when two or three Topology Actions (Tas) start to be considered, and it is estimated that this number of TAs alone could add a further 1–3% of savings, complex NTO aside.  For NESO, this can be equivalent to £75m per year in cost reduction when considered against current balancing costs. 
 
This project demonstrated that NTO is not a commercial off-the-shelf product, meaning NTO cannot have a ‘plug-in’ implementation. Rather, solutions require a level of customisation to cater individual TSO business practices, but the core solution features can be shared across multiple TSOs. NTO goes far beyond system design and implementation. It requires reimagining control room processes to standardise and optimise control room operations. This project demonstrated the importance of an expert team that understands the linkage between technology and business will add significant value when developing a workable NTO solution. Designing an NTO solution requires a multidisciplinary team with expertise in power systems, data engineering, network modelling, data science, and programme management. This will ensure a technically skilled team can manage and coordinate the implementation end-to-end, and that relevant departments within NESO & wider global community are active participants. 

Lessons Learnt

Implementing the NTO programme over a 10-year horizon represents a complex transformation initiative that demands substantial expertise and resources. Lessons Learnt on an approach that best release value up to March 2028 is: 
> NTO is not a commercialoff the shelf product available for implementation, meaning NTO cannot have a ‘plug-in’ implementation. Rather, solutions require a level of customisation to cater to individual TSO business practices, but the core solution features can be shared across multiple TSOs. 
> Based on the TSOs interviewed & desktop research conducted, a full NTO solution addressing all types of constraints including thermal, voltage and dynamic stability implemented as a decision support tool does not exist. 
> Vendors have developed linear mathematical formulations addressing mainly thermal constraints, areas of voltage and dynamic stability congestion management are largely unexplored. These unexplored areas must be further developed to ensure a full workable NTO solution is available for control room applications. This includes formulation of fast and full AC solvers, which are essential for capturing these complexities.

> NTO solutions have been formulated using a single mathematical formulation, using Mixed Integer Programmes (MIP). However, MIP ultimately accommodates the continuous and discrete variable types seen in NTO, only in a manner that increases computational intensity. Modern developments in solutions have focused on dual formulations of Topology Actions (TA) and power flow as enables more scalability, customisation and streamlining of methods. 
> Limited number of Topology Actions (TA) have been incorporated in NTO formulations to date, future work should expand these techniques to include a wider range of simultaneous TAs such as busbar couplers, switches, reactive power devices, dynamic line ratings, VAR compensators, tripping schemes, and active network management. 
> Vendors and TSOs have primarily developed solvers for CPU-based environments. Tests on large-scale networks (e.g. >10,000 buses) reveal underutilisation of available CPU resources. To improve performance, future implementations should incorporate multi-threading, multi-processing, and full workload distribution across cores. 
> GPUs grant faster computational; speed leading to more Topology Actions (TA) sets to be optimised in less time. Future work should seek to solve AC PF with a linear treatment, or a partial linear-nonlinear treatment where GPUs and CPUs work in parallel. 
> Vendors and TSOs have mainly focused on development of UI tools for control support applications. Full automated UI and UX deployments are in the early stages, and a wide gap has been observed for usability among control room operators. 
> Vendors and TSOs have not fully integrated NTO solutions into existing IT and OT platforms. Automated data orchestration are in the early stages. 
> Data and model alignment remain an important enabler for scaled deployment of NTO into control room as a workable decision support tool. Input data and quality of model alignment can impact significantly on the accuracy of the output of NTO solutions.