SILVERSMITH will investigate compliance issues that are expected to grow on our Low Voltage network as Low Carbon Technologies (LCT) and Distributed Generation (DG) connections increase. Phase 1 will conduct two network studies, the first using the Transform tool, the second using PowerFactory DIgSILENT. We will assess the demographic of future issues and produce an opportunity statement for devices to resolve them after evaluating novel technologies. Phase 2 then re-assess the network model using these technologies and produce a list of functional requirements and design methodology to selecting the best technology would be under different network conditions. To conclude, a comparison between each approach will be made to investigate what method of network analysis leads to the best investment decision making.
Benefits
There has not been a vendor independent review into voltage control technologies on the Low-Voltage (LV) networks. At this pre-adoption stage, we believe it is important to fully understand the scale of benefit available, rather than progress straight to a large-scale demonstration or trial of a particular technology. Project SILVERSMITH involves a bottom-up assessment to learn what functional requirements are needed to resolve future LV network issues, which may arise from proliferated low carbon technology connection. The analysis will estimate the demographic these issues over our license area and evaluate the state-of-the art in a comprehensive literature review. As investment planning on the LV typically has been fairly limited, we want to compare what methods are most useful. Therefore, we will carry out our analysis in two parallel streams. Our first uses the Transform tool that has been used by DNOs for network investment planning. The second will study a series of representative networks on ConnectLV and DIgSILENT and focus on more in-depth power flow analysis. Using the results of this work, DNOs will be able to coordinate their LV voltage control investment to tackle the most pertinent issues that are expected to arise at LV. Furthermore, by comparing the conclusions made by each approach, DNOs will be able to understand which network modelling techniques is better suited for investment decision making.
Learnings
Outcomes
The outcomes of the project are as follows:
- A detailed literature review which documents the state-of-the-art technologies which are being developed by suppliers.
- A breadth-based Network Study which uses the economic model Transform to estimate the demographic of future network constraints witnessed on the LV network from now up until 2050. Different DFES scenarios consider the impact of the LCT uptake rate on network constraints.
- A depth-based Network Study using DIgSILENT’s PowerFactory software to evaluate three case study networks. After connecting LCTs according to the three DFES scenarios, analysis was performed in study years: 2028, 2033, 2040, and 2050.
- A report detailing the Functional Requirements expected to be needed by each archetype was produced. This details the range of BAU technologies which would be used today, and also where novel technologies are used in place of BAU approaches, detailing the financial savings which could be possible.
- A power-flow-based study evaluated each of the novel technologies considered in our project. This study evaluated the relative strengths and weaknesses of the technologies.
- A comprehensive LV Voltage Control Selection Methodology was produced, which clearly explained the types of technologies that should be considered for addressing network compliance issues in different case studies. It also highlighted the scale of customer flexibility which would be required to defer and avoid BAU reinforcement.
- A summary report, that gives an overview of how each technology is applied to our network in the study, and reasons why some of the novel technologies were not selected.
- The information contained within the SILVERSMITH reports provides benefit to our customers, by allowing us to direct our innovation to the most critical parts of our network requirements. It focuses the effort for future innovation, and provides a stimulus for supply chain engagement.
All reports can be found on the SILVERSMITH project page: https://www.nationalgrid.co.uk/innovation/projects/solving-intelligent-lv-evaluating-responsive-smart-management-to-increase-total-headroom-silversmith
Lessons Learnt
State-of-the-art LV Voltage Control devices
- As part of the literature review, a range of solutions have been considered that have the potential to provide either additional voltage or thermal capacity to the network. These solutions range from retrofit devices, physical network interventions, market solutions (encouraging consumer engagement) and policy solutions.
- The Literature Review documents the operating parameters of different LV voltage control devices.
Modelling LCT uptake using Distribution Future Energy Scenarios (DFES)
- Currently, no commercial heat pump profile exists. Within this project, this limited the insights and results obtained for LV feeders, which had more commercial customers connected.
- It was recommended that an understanding of commercial heat pump profile(s) should be developed such that the effect of commercial heat pumps can be included in future network modelling.
Power flow models
- During the PowerFactory network modelling, several delays were caused by a lack of model convergence. LV network studies are not currently performed as a business-as-usual activity, as such the electrical diagrams are often incomplete or missing detail. Where information was missing, approximations and assumptions were made, which often caused a lack of convergence until improvements were made incrementally.
- In future work, it is recommended to err on the side of caution when planning detailed network modelling using LV records. It is also recommended that to improve this process, systematic improvements to the LV records should be made.
- The networks modelled in the PowerFactory work at the project's outset were limited in their ability to evaluate novel solutions. D1.1 showed that very few issues were expected to arise looking out to 2050. When it came to assessing the functional requirements of the novel options, solutions were limited to improving areas of the network that didn’t experience significant issues. In particular, remote switching/meshing was only possible to be modelled on one area of the urban model.
- One of the main objectives was to understand how often LCT loads would require certain proactive measures to be taken on the LV network. For instance, we wanted to understand how increased LCT loading and export at different times throughout the day may require multiple interventions throughout the day to manage the voltage. By only investigating three scenarios, we were not able to accurately model the impact of this across the network and understand where this may be needed at scale.
The demographic of network issues
- Feeder types dominated by commercial customers are forecast to witness primarily voltage rise constraints. This is due to the uptake of PV.
- Transformer load-related constraints are consistently the most significant issue across all licence areas.
- In 2028, on average across all National Grid’s licence areas approximately 6% of feeders require intervention for thermal constraints under net import (approximately 4.5% transformer and 1.5% cable), and 2% of feeders require intervention for voltage rise constraints.
- By 2040, thermal constraints will affect 40% of feeders. 27% Transformer, 13% thermal. This increases to 60% by 2050.
- While commonly a single network constraint is encountered, it is also common that multiple types of network constraints are encountered at the same time. When multiple network constraints are encountered, these are sometimes on distinct feeders, but regularly also on the same feeder. A common example would be thermal transformer and thermal cable constraints being witnessed simultaneously on a feeder.
- The network results showed that the Dense Urban network is more likely to experience breaches in thermal capacity significantly before statutory voltage limits are breached. This may mean that solutions which may become necessary for the Dense Urban network will favour those which can reduce circuit loading, particularly during the winter evening peak periods. Flexibility may be a strong solution in Dense Urban networks, which are estimated to require around 1.0 kW of demand turn-down per customer.
- In the Urban Network, higher levels of PV generation and longer feeders, when compared to the Dense Urban network, show that statutory upper voltage limits (+10%) could begin to be breached from 2033. It was also found that thermal loading begins to become overloaded during the winter peak from 2033 (Leading the Way) and 2040 (Best View) at which point some intervention would be necessary. For the Urban Network, interventions which reduce the thermal loading during the winter whilst also improving voltage management in the summer will be necessary.
- For the Rural Network, some customers will experience high voltages breaching the upper statutory voltage limits during peak PV export from 2040 onwards. Although circuit loadings remain well below their 100% rating the total loading on the supply transformer will begin to be exceeded for the highest LCT uptake scenarios (Leading the Way). In some situations, circuit loading remains within thermal limits but in Rural networks, voltage issues may begin to occur at different points along the feeder.
Performance of Novel LV Voltage Control Devices
- To use a STATCOM successfully, remote monitoring of the network is needed. This is to feed into an algorithm which triggers when the STATCOM needs to be operated. Failing to do so would increase losses and consumer capacity.
- Ideally, a STATCOM would only resolve voltage issues during the middle of the day when the voltage rise is the highest due to PV. At other times of the day, it was found to consume too much of the network's capacity due to the additional reactive power flow.
- Harmonic filtering and phase balancing offer benefits which primarily apply to the transformer and the HV network upstream of the connection point. On the LV network, harmonic currents will still exist as the active injection to cancel the harmonics will only benefit the network upstream of the filter.
Most commonly used technologies
The novel solutions most commonly deployed across National Grid’s LV network and therefore offering value over business-as-usual solutions are:
- Network data monitoring
- Active network management (dynamic control of the network by controlling, for example, normally open points)
- Active transformer cooling
- Real-Time Thermal Ratings for HV/LV Transformers
- If the ENA140 Consultation which is considering whether to widen voltage tolerances is passed, manual tapping to increase headroom could be accommodated on networks where previously it would not be deployable due to voltages dropping below the statutory voltage minimum limit.
Sensitivity studies
- Similar solutions are selected to be deployed across all four of National Grid’s electricity distribution licence areas. The proportion of solution deployment is sensitive to the LCT uptake rates on each licence’s area feeder set. Any solutions deployed should be based on the network constraint type and extent of constraint witnessed on the specific feeder.
- The number and types of distinct solutions required across National’s Grid network are independent of the scenario. Regardless of the actual LCT uptake rate, distinct solutions will be utilised for particular feeders due to the variability in the clustering of LCT deployments. Network Operators do not need to be concerned about the uptake rate of LCT across the system, instead, they should focus on the constraint type and extent caused by LCTs on each feeder.
Consideration of flexibility
- The use of domestic flexibility (technology agnostic) was considered as part of the Selection Methodology. The average amount of flexibility per feeder and per customer on that feeder was calculated to i) defer the initial reinforcement by 5 years, and ii) remove the need for the second wave of reinforcement. Based on these studies, to defer network reinforcement for 5 years from the time of the first network constraint, a change in demand at system peak of between 0.7kW and 2.1kW per customer is required depending upon the network archetype.
- The cost and duration of this flexibility was beyond the scope of this work, but is being considered in our NIC project EQUINOX which this learning has supported.
- As further trials investigating domestic flexibility are conducted, more accurate information will be available to estimate the amount of demand side response available to shift demand away from peak times, and the expected cost which DSOs would be willing to pay.
Opportunity for future innovation
- It was estimated that solutions offering cable headroom increased between 15% and 100%, for less than £50,000, would have significant value. Where a network experiences a cable headroom requirement of between 15% and 50% headroom uplift, Transform would pick network meshing. However, for practical reasons, it is unlikely that retrofitting classic Network Meshing would be achievable at scale which leaves a significant opportunity. In cases where network meshing is not possible, the next logical solution is splitting the feeder at a totex cost of £53,880. Therefore, solutions that can release greater than 15% thermal cable capacity with less cost and technical challenge than that of permanent network meshing would offer significant value.
- Innovative technologies that could offer between 20% and 80% thermal transformer capacity at less than approximately £16,000 per feeder (for feeders supplied by GMTs) or £7,500 per feeder (for feeders supplied by PMTs) would provide significant value to the network operator. Innovation activity could investigate solutions able to fill this gap. Appendix 2 provides details of each network archetype including the average number of feeders per archetype, which varies between 1 and 5.
- The available solution set has a gap between cheaper minor improvements and costly major improvements. Where solutions offering minor improvements don’t address the network issues, we are forced to deploy far more costly solutions. Intermediate solutions would offer significant savings.
Recommended technology developments
- Of the available technologies, active transformer cooling was widely selected and recommended for a trial by the late 2020s.
- By the early 2030s, trial Real-Time Thermal Ratings for HV/LV Transformers across a sample of National Grid’s pole-mounted transformers are forecast to become thermally constrained. Trials should focus on how to install equipment necessary for RTTR on PMTs and on quantifying the benefit of RTTR on thermal PMT capacity.
- By the mid-2030s trial active network management across a section of National Grid’s LV network and quantify the thermal and voltage headroom release achieved.
- Trial new use cases of network data monitoring. This includes real time flexibility forecasting, real time thermal ratings, and temporary ratings to increase the utilisation of existing assets.