Automation in network modelling: Freeing planners to plan, strengthening the data foundations of the future grid
At a glance
Automation can give electricity network planners more time to focus on the decisions that need their engineering judgement. Building reliable models starts with stronger data foundations and clear ownership of information. When those foundations are in place, repeatable processes can reduce preparation time, support faster connection assessments and help planners respond to growing demands on the grid.
Explore how automation in network modelling can help electricity network planners save time, strengthen data foundations and support faster grid connection assessments.
Electricity network planners face rising workloads as demand shifts, new generation connects and existing infrastructure requires reinforcement. Making the best use of their expertise depends on reducing the time spent preparing and maintaining the information that supports their decisions.Electricity network planners face rising workloads as demand shifts, new generation connects and existing infrastructure requires reinforcement. Making the best use of their expertise depends on reducing the time spent preparing and maintaining the information that supports their decisions.
GHD’s GRID roundtable, Automation in Network Modelling: Freeing Planners to Plan, explored how to achieve this. Drawing on work with National Grid Electricity Distribution (NGED), Neil Murdoch and Nick Edwards outlined the challenge of combining existing data sources into reliable network models.
Speaking afterwards, GHD energy consultant Thomas Bennett explained the delivery implications. Residential networks with historically stable consumption now face electric vehicle charging, new generation and housing development. With trained engineers in short supply, reducing repetitive preparation gives them more time to assess networks and plan improvements.
GHD’s GRID roundtable, Automation in Network Modelling: Freeing Planners to Plan, explored how to achieve this. Drawing on work with National Grid Electricity Distribution (NGED), Neil Murdoch and Nick Edwards outlined the challenge of combining existing data sources into reliable network models.
Speaking afterwards, GHD energy consultant Thomas Bennett explained the delivery implications. Residential networks with historically stable consumption now face electric vehicle charging, new generation and housing development. With trained engineers in short supply, reducing repetitive preparation gives them more time to assess networks and plan improvements.
The Jenga problem
A Jenga analogy raised by Cherryanne from Siemens captured how successive layers of data and systems have accumulated over time, leaving their stability dependent on the foundations beneath them.
Network information has evolved from paper records into geographic information systems (GIS), asset databases and control systems. Each serves a purpose, but they were not necessarily designed to work together or support automated modelling. Asset locations, electrical characteristics and switching arrangements may sit in separate systems under different identifiers.
Bennett highlighted another difficulty: information can be replicated without a clearly defined authoritative source. When records are updated inconsistently, engineers must establish which reflects reality. In extreme cases, this can require another physical inspection.
His recommendation is to establish an authoritative source for each piece of information, with clear processes governing updates and resolving conflicts, even where legacy systems remain.
Network information has evolved from paper records into geographic information systems (GIS), asset databases and control systems. Each serves a purpose, but they were not necessarily designed to work together or support automated modelling. Asset locations, electrical characteristics and switching arrangements may sit in separate systems under different identifiers.
Bennett highlighted another difficulty: information can be replicated without a clearly defined authoritative source. When records are updated inconsistently, engineers must establish which reflects reality. In extreme cases, this can require another physical inspection.
His recommendation is to establish an authoritative source for each piece of information, with clear processes governing updates and resolving conflicts, even where legacy systems remain.
Connecting the dots
The NGED project involved replacing a legacy modelling tool and automating model production for its 11kV network. Making the information usable required close collaboration with planners and data owners.
The team combined geographic, asset and control-system information, encountering inconsistent identifiers, missing references and connectivity gaps. A break in a feeder’s recorded connectivity could leave downstream assets disconnected in the model; missing identifiers could exclude assets altogether.
Transformation rules, engineering judgement and corrections to source information addressed these issues through repeated testing. Broader sampling was also essential: successful initial models did not guarantee consistent data elsewhere, even within the same organisation.
The team combined geographic, asset and control-system information, encountering inconsistent identifiers, missing references and connectivity gaps. A break in a feeder’s recorded connectivity could leave downstream assets disconnected in the model; missing identifiers could exclude assets altogether.
Transformation rules, engineering judgement and corrections to source information addressed these issues through repeated testing. Broader sampling was also essential: successful initial models did not guarantee consistent data elsewhere, even within the same organisation.
Turning data checks into action
Identifying poor data only adds value when teams can act on it. Participants noted the limitations of high-level quality scores that flag problems without showing their location or how to resolve them.
Targeted checks could identify discrepancies between expected demand and the load represented in a model, directing teams to connectivity issues. Feeding corrections back into source systems improved later models and reduced repeated effort.
Data improvement therefore became an ongoing process involving those who create, maintain and use the information. Building these working relationships was as important as the technical tools.
Targeted checks could identify discrepancies between expected demand and the load represented in a model, directing teams to connectivity issues. Feeding corrections back into source systems improved later models and reduced repeated effort.
Data improvement therefore became an ongoing process involving those who create, maintain and use the information. Building these working relationships was as important as the technical tools.
Compounding benefits
More consistent models free engineers to focus on analysis. The team described automated studies of demand growth, potential overloads and voltage constraints, helping identify where further investigation or investment may be required.
Bennett suggested the greatest benefits may come from frequently repeated activities, such as connection assessments. Starting each study with an up-to-date model reduces preparation time and lets requests be processed faster.
The same information supports longer-term planning, extending the value of improvements across daily operations and strategic investment decisions. Automation enables engineers to examine more scenarios, while their judgement remains essential in deciding how to resolve constraints.
Bennett suggested the greatest benefits may come from frequently repeated activities, such as connection assessments. Starting each study with an up-to-date model reduces preparation time and lets requests be processed faster.
The same information supports longer-term planning, extending the value of improvements across daily operations and strategic investment decisions. Automation enables engineers to examine more scenarios, while their judgement remains essential in deciding how to resolve constraints.
Keeping the process transparent
The roundtable also explored the potential for AI to help cleanse data, identify relationships and connect disparate information sources.
Participants emphasised the importance of retaining visibility from inputs to outputs. Engineers need to understand the assumptions behind a model, test their implications and explain why a result is credible.
Where complete information is unavailable, assumptions may be necessary. Recording them consistently and assessing their sensitivity helps teams understand the limits of the analysis and compare results over time.
Trust also depends on adoption. Moving to a new modelling environment requires training, support and opportunities for users to build confidence. As familiarity grows, returning to training can help teams use more of a tool’s capabilities.
Participants emphasised the importance of retaining visibility from inputs to outputs. Engineers need to understand the assumptions behind a model, test their implications and explain why a result is credible.
Where complete information is unavailable, assumptions may be necessary. Recording them consistently and assessing their sensitivity helps teams understand the limits of the analysis and compare results over time.
Trust also depends on adoption. Moving to a new modelling environment requires training, support and opportunities for users to build confidence. As familiarity grows, returning to training can help teams use more of a tool’s capabilities.
Better data, better outcomes
The roundtable demonstrated the work needed to establish reliable models: integrating information, correcting inconsistencies and developing processes that keep the data usable over time.
Bennett’s perspective connects the technical work to the wider delivery challenge. Engineers have developed specialist skills over many years. Reducing repetitive preparation allows them to apply those skills to constraints, options and investment decisions.
Reliable data and repeatable processes strengthen every study, helping planners identify capacity constraints, compare options and build the case for investment. As workloads grow, this supports faster connection decisions and clearer priorities for network reinforcement.
Bennett’s perspective connects the technical work to the wider delivery challenge. Engineers have developed specialist skills over many years. Reducing repetitive preparation allows them to apply those skills to constraints, options and investment decisions.
Reliable data and repeatable processes strengthen every study, helping planners identify capacity constraints, compare options and build the case for investment. As workloads grow, this supports faster connection decisions and clearer priorities for network reinforcement.